An integrated artificial neural network/polymer-based pH sensor: a new engineering perspective to conducting polymer technology

A new approach towards the development of a novel pH sensor using electrically conducting polymers is addressed. The approach is based on computer analysis of different experimental works in which various conducting polymer-based pH electrodes are examined. In this work, we present a novel conducting polymer pH sensor by which the pH in solutions with different pH values can be detected. Since re-usability and reproducibility of conducting polymer sensors have been themselves controversial issues, we discuss the application of artificial intelligence (AI) methods, artificial neural networks (ANNs) in particular, to address and examine such an issue on the performance of the introduced pH sensor. The introduced integrated ANN/conducting polymer pH sensing system has shown much better reproducibility in detecting pH of different acidic and basic media over the conventional polymeric pH detectors. The system can be designed for pH detection in any operational environment in an online/real time manner. Having access to such a system enables us to have access to the history of a desirable chemical process with respect to the extent of the acidity of the process.     

Connectionist learning: A comparison of neural networks and an optical thin-film multilayer model

Current work on connectionist models has been focused largely on artificial neural networks that are inspired by the networks of biological neurons in the human brain. However, there are also other connectionistarchitectures that differ significantly from this biological exemplar. We proposed a novel connectionist learning architecture inspired by the physics associated with optical coatings of multiple layers of thin-films in a previous paper (Li and Purvis 1999, Annals of Mathematics and Artificial Intelligence, 26: 1-4). The proposed model differs significantly from the widely used neuron-inspired models. With thin-film layer thicknesses serving as adjustable parameters (as compared with connection weights in a neural network) for the learning system, the optical thin-film multilayer model (OTFM) is capable of approximating virtually any kind of highly nonlinear mappings. The OTFM is not a physical implementation using optical devices. Instead, it is proposed as a new connectionist learning architecture with its distinct optical properties as compared with neural networks. In this paper we focus on a detailed comparison of neural networks and the OTFM (Li 2001, Proceedings ofINNS-IEEE International Joint Conference on Neural Networks, Washington, DC, pp. 1727-1732). We describe the architecture of the OTFM and show how it can be viewed as a connectionist learning model. We then present experimental results on solving a classification problem and a time series prediction problem that are typical of conventional connectionist architectures to demonstrate the OTFM's learning capability.     

Quantification of the capability of micro-CT to detect defects in castings using a new test piece and a voxel-based comparison method

We present a modified aluminium casting which is especially suited as test piece for measuring casting defects and the geometry by means of cone-beam micro-focus X-ray systems, and which may become a reference standard for dimensional measurements and defect detection.To obtain a test piece with inner geometries measured by tactile means, we divided a small aluminium cylinder head into four pieces in such a way that most inner surfaces can be reached with a tactile probe. Reference geometries (spheres and cylinders) were applied to define a coordinate system for aligning the measurements in the disassembled and re-assembled state. The four pieces were re-assembled after the tactile measurement.The test piece also contains casting defects. In order to be able to use the assembled cylinder head as reference sample for defect detection, measurements with higher spatial resolution and better signal-to-noise ratio were performed on the single parts. For improving the reliability of the reference measurements, CT measurements of each part were carried out in three different orientations, and the individual defect detections were combined to obtain a reference data set with a high probability of defect detection and a low rate of erroneous detections.A new method for comparing the defect detection in a CT measurement to a reference data set is demonstrated, which provides individual information on every detected flaw.We discuss the results of measurements in the assembled state with respect to the reference data for flaw detection.     

Improving harmony search algorithms by using tonal variation: the case of Sudoku and MKP

In this paper we propose an improved Harmony Search (HS) version inspired in the tonal variation of jazz musical improvisation. To evaluate our approach we considered two well-known problems, a Constraint Satisfaction Problem: Sudoku, and a Constraint Satisfaction Optimisation Problem: the Multidimensional Knapsack Problem (MKP). For each problem, we considered an existing baseline HS algorithm to implement our technique: the HS for Sudoku puzzles and, the Adaptive Binary HS for the MKP. The experiments showed that including tonal variation allows HS algorithms to find better quality solutions in both problems.     

一种新型神经网络学习控制器及其在气动压力控制系统中的应用

本文利用神经网络非线性映射的特点,用逆向建模的思想,设计了一种基于神经网络的学习控制器,并将其应用于气动压力控制系统,收到了很好的效果。     

Greedy information acquisition algorithm: A new information theoretic approach to dynamic information acquisition in neural networks

In this paper, we proopose a new information theoretic approach to competitive learning. The new approach is called greedy information acquisition , because networks try to absorb as much information as possible in every stage of learning. In the first phase, with minimum network architecture for realizing competition, information is maximized. In the second phase, a new unit is added, and thereby information is again increased as much as possible. This proceess continues until no more increase in information is possible. Through greedy information maximization, different sets of important features in input patterns can be cumulatively discovered in successive stages. We applied our approach to three problems: a dipole problem; a language classification problem; and a phonological feature detection problem. Experimental results confirmed that information maximization can be repeatedly applied and that different features in input patterns are gradually discovered. We also compared our method with conventional competitive learning and multivariate analysis. The experimental results confirmed that our new method can detect salient features in input patterns more clearly than the other methods.     

宝钢1550mm冷轧生产线新技术介绍

介绍了宝钢三期工程1550mm冷轧生产线的概况,并分别介绍了酸洗－冷连轧、连续退火、连续热镀锌、连续电镀锌、电工钢连续退火涂层等主要生产机组采用的先进技术和主要技术参数．     

基于神经网络和遗传算法的冲压成形多目标优化技术

在冲压成形质量控制中,目标质量间常常是相互冲突的(如破裂和起皱)。传统求解多目标优化问题取决于设计人员对优化模型的理解程度、实践经验等,求解的结果在工程中并非为最合理。文章提出一种集数字化分析技术、神经网络和遗传算法于一体的冲压成形多目标优化设计技术。其以数字化分析的大量结果作为神经网络的学习样本,遗传算法所需的目标函数值由神经网络模型预测,该技术实现了多目标优化过程中遗传算法个体适应度值的动态求解,从而解决了数字化分析计算量大的缺陷。实例验证了该优化技术的有效性,为冲压成形优化设计提供了一种新的方法。     

Tantalum protective thin coating techniques for the Chemical Process Industry: Molten salts electrocoating as a new alternative

A comparison of corrosion resistance and basic properties of solid tantalum with other high-performance materials used in the Chemical Process Industry (CPI) is given. The corrosive chemicals taken into consideration are strong acidic media. Secondly, it is pointed out that tantalum, which exhibits excellent corrosion resistance, owing to a rapid build-up of passivating protective film in oxidizing conditions, also has good mechanical, thermal and electrical properties which suggest its use when little or no metallic corrosion is tolerated. Thirdly, tantalum thin-layer, coated onto a usual base metal, which offers the same protection as solid metal and avoids its expensive use, is treated. Fourthly, numerous tantalum-coating techniques for clad-vessel and CPI devices are reviewed and compared. Amongst these coating techniques, this paper focuses mainly on two techniques which give a very thin, protective coating against corrosion. Thus, Chemical Vapor Deposition (CVD) and Molten Salt Electro-deposition (MSE) are especially enhanced. Finally, MSE which is still not widely used for manufacturing clad-vessels is examined in greater detail.     

Study of two new aqueous alkaline Copper cleaning solutions and comparison with a conventional bath

This paper studies the cleaning of commercial copper with two new aqueous alkaline solutions containing formaldehyde and D(+)glucose at room temperature. The results are compared with a conventional cleaning solution. Three organic corrosion inhibitors were used: Phloroglucinol, 8-Quinolinol and Resorcinol. Rest potential, polarization resistance and polarization curves were the electrochemical techniques used. The gravimetric technique was also used. It has been observed that the two new cleaning solutions remove copper tarnishing. The solution with D(+)glucose produces the best surface finish. Likewise, the conventional solution is an excellent copper cleaning bath. The order of inhibition efficiency of the organic inhibitors is as follows: 8-Quinolinol > Phloroglucionol > Resorcinol.     

A comparison of the pulsed,lock-in and frequency modulated thermography nondestructive evaluation techniques

Pulsed, lock-in and frequency modulated thermography are three alternative nondestructive evaluation techniques. The defect imaging performance of these techniques are compared using: matched excitation energy; the same carbon fiber composite test piece and infrared camera system. The lock-in technique suffers from “blind frequencies” at which phase images for some defects disappear. It is shown that this problem can be overcome by using frequency modulated (chirp) excitation and an image fusion algorithm is presented that enhance phase imaging of defects. The signal-to-noise ratios (SNRs) of defect images obtained by the three techniques are presented. For the shallowest defects (depths 0.25 and 0.5 mm, 6 mm diameter), the pulsed technique exhibits the highest SNRs. For deeper defects the SNRs of the three techniques are similar in magnitude under matched excitation energy condition.     

利西拉来治疗2型糖尿病的有效性及安全性的Meta分析

目的: 系统评价利西拉来与安慰剂比较治疗2型糖尿病的疗效和安全性。方法: 计算机检索PubMed、EMBASE、OVID、Cochrane、CNKI、VIP、CBM数据库,查找利西拉来与安慰剂比较治疗2型糖尿病的随机对照试验(randomized controlled trial,RCT),对纳入研究进行方法学质量评价,并用RevMan 5.2 软件进行Meta分析。通过Stata 12.0 软件对其是否对称按Begg's法进行分析,另外通过考察单项研究对总合并效应量的影响来进行敏感性分析,以评价结论的稳健性。结果: 共纳入8个RCT,随访时间≤24周的Meta分析结果显示:利西拉来在控制糖化血红蛋白(hemoglobin A1C,HbA1c)＜7%[RR=1.96,95%CI(1.65,2.32),P＜0.00001,Z=7.79]或≤6.5%[RR=2.70,95%CI(2.26,3.23),P＜0.00001,Z=10.83]的达标率方面均优于安慰剂组。总不良反应及低血糖、注射部位反应、胃肠道反应、恶心、呕吐、头晕的发生率与安慰剂组相比,利西拉来组均高于安慰剂。头痛、腹泻的发生率两组相似,无统计学差异。所有不良反应均较轻微,绝大多数患者可以耐受。随访时间为76周时,Meta分析结果显示:利西拉来在控制糖化血红蛋白＜7%[ RR=1.30,95%CI(1.12,1.51),P=0.0005,Z=3.49]或≤6.5%[ RR=1.41,95%CI(1.12,1.78),P=0.004,Z=2.89]的达标率方面均优于安慰剂组。总不良反应、注射部位反应、胃肠道反应、恶心、呕吐的发生率与安慰剂组相比,利西拉来组均高于安慰剂,而低血糖、腹泻两组无明显差异。结论: 利西拉来可以有效控制2型糖尿病患者血糖,利西拉来主要的不良反应是低血糖、注射部位反应、胃肠道反应、恶心、呕吐、头晕、头痛、腹泻,但不良反应较轻微,因此可以用来治疗2型糖尿病,但远期结果有待高质量、大样本、长期随访的临床研究予以验证。     

一种新型浮式综合发电平台的概念设计及其动态响应研究

参照OC3-Hywind Spar Buoy和1TI Energy Barge(以下分别简称Spar和Barge)的结构特点,提出了一种新型浮式综合发电平台的概念,根据海洋能发电的特性,构建了“风能-太阳能-波浪能”的联合发电系统,以实现多种能源的耦合互补,并以我国南海作为平台工作条件,采用ANSYS的水动力学模块对平台进行了数值计算.结果表明:该平台的动态响应较小,水动力性能良好,适合在南海海域使用.     

Competition between cue response and place response: a model of rat navigation behaviour

Different neural systems are involved in animal navigation depending on the type of task. Experimental studies support the idea that the hippocampus is necessary to learn a spatial representation required to navigate toward hidden goals (place response), whereas the dorsolateral striatum is involved in the learning of stimulus-response associations when navigating toward visible (or cued) goals. These systems compete for action selection according to the characteristics of the task, previous experience (e.g. raining procedure) or endogenous factors. This paper reviews both experimental data on the theory of multiple memory systems involved in navigation, and a recent computational model of action selection based on the competition of place and cue responses learnt during training. The model implements separately the two types of response, i.e. place response and stimulus response. Furthermore, competition takes place to select which behaviour will actually be performed. The model was tested in a simulated environment using a protocol analogous to those used in experiments with animals.     

Removal and repair techniques for thermal barrier coatings: a review

ABSTRACT

A comprehensive literature review of the existing techniques for removing and repairing of damaged thermal barrier coatings is presented, with the focus on top ceramic coats. The advantages and disadvantages of each technique are compared and assessed. The review shows that there is not a universal method applicable to all coating systems. The selection of the coating removal and repair process must be specific to damaged coating systems, based on their composition, type of damages, and available resources. This review will provide some inside look at various approaches in an effort to meet the different coating repair needs.     

信息革命:催生全新的互联网精神

<正>十多年之前谈起互联网,大多数人恐怕认为这只是一种新的技术、新的服务,属于那些学生和年轻人的游戏而已。但是时至今日,再看我们身处的这个世界,发现由互联网掀起的这场持久的信息革命,已大大超越了技术和游戏的范畴,越来越多地影响、改变,甚至颠覆着各种行业规则与组织方式,互联网孕育了许多的暴富神话、传奇人物、数不胜数的极致产品,以及眼花缭乱的创新模式。无论是人物,产品还是模式,内在都体现出一种独到的精神特质,我们称之为互联网精神。它具体可以概括为四个方面:将产品做成一种服务,打破信息壁垒做交易,发展生态化的组织管理,以及敢于跨界的创新。     

Merging subsets of attributes to improve a hybrid consistency-based filter: a case of study in product unit neural networks

This paper presents a quality enhancement of the selected features by a hybrid filter-based jointly on feature ranking and feature subset selection (FR-FSS) using a consistency-based measure via merging new features which are obtained applying other FR-FSS evaluated with a correlation metric. The goal is to overcome the accuracy of a neural network classifier containing product units as hidden nodes combined with a feature selection pre-processing step by means of a single consistency-based FR-FSS filter. Neural models are trained with a refined evolutionary programming approach called two-stage evolutionary algorithm. The experimentation has been carried out in eight complex classification problems, seven out of them from UCI (University of California at Irvine) repository and one real-world problem, with high test error rates (around 20%) with powerful classifiers such as 1-nearest neighbour or C4.5. Non-parametric statistical tests revealed that the new proposal significantly improves the accuracy of the neural models.     

Disorientations as a function of distance: a new procedure to analyze local orientation data

A new procedure for analyzing local orientation data has been developed to investigate the relation between orientations and data point distances. The key idea is to determine the distribution of the disorientation angle as a function of the distance between the data points. The evolution of the disorientation angle distribution with increasing data point distance provides quantitative information about average local variation of the crystal orientation which can be used to classify structural elements, to determine the average structural size, and allows the quantification (crystallographically) of the boundaries between adjacent structural elements. The potential of this analyzing procedure is demonstrated by investigating the orientation data of Electron Backscattering Diffraction (EBSD)-measurements performed on severe deformed copper samples. The benefits in comparison to other commonly used methods to quantify local disorientations are discussed.     

Application of artificial neural networks to the real-time operation of conducting polymer sensors: a pattern recognition approach

An artificial neural network (ANN)-based pattern recognition method is adopted for conducting polymer (CP) sensors. The method is capable of creating different patterns and models based on an on-line data collection from a multichannel analog/digital (AD) device. The flow of information is directed from the surface of the CP electrode into an AD device which is connected to an ANN-trained computer. The ANN software (Turbo Neuron) used in this study accepts the data as its inputs and creates the best possible patterns, based on pre-selected parameters, to classify the type of ions existing in the operational environment. The method is recommended to be used in the field of CPs where passive analytical methods have not been successful in addressing reusability of the CP electrodes.