APPLICATION OF MAXIMUM ENTROPY METHOD TO CALCULATE RDF OF Ni-P AMORPHOUS ALLOY FROM X-RAY DIFFRACTION DATA

Experiments clarified that application of the maximum entropy method to calculate the RDF(radial distribution function)of the Ni-P amorphous alloy from fewer data of its X-raydiffraction data shows a higher resolution and good consistent with known results.It is be-lieved that this method is available for the analysis of amorphous structure.     

熔体快淬La1-xNdxFe11．5Si1．5（x=0，0．3，0.5）化合物磁性能和磁熵变

利用熔体快淬法制备了La1-xNdxFe11．5Si1．5（x=0，0．3，0.5）快淬带。通过X射线衍射和振动样品磁强计测试了回火态（1000℃，5h）La1-xNdxFe11．5Si1．5）快淬带的相组成和磁性能。结果表明：不同成分快淬带回火后均可得到稳定的NaZnl，型化合物。和常规的熔炼法相比，用熔体快淬法通过短时间的回火（1000℃，5h）可以制成具有大磁熵变的NaZn13型的化合物。用Nd替代部分的La后，居里温度略有提高，同时保持有大的磁熵变｜△S｜；并且随着Nd含量的增加，有效制冷温区有变宽的趋势。     

基于熵权理论和渴求函数法的多响应优化设计研究

文章首先对多响应优化问题进行了回顾和分析,针对多响应优化设计中权重确定和优化模型的建立问题,探讨了熵权理论和基于过程能力指数的渴求函数法在多响应优化设计中应用的可能性,并提出了熵权理论和渴求函数法结合的多响应优化设计方法,然后通过实例对文中所提的方法进行了实证研究,最后得出结论通过将熵权理论和渴求函数结合应用于多响应优化问题中,可以达到很好的效果。     

Effect of welding speed and electrode extension on the approximate entropy of welding current in short-circuiting GMAW

Based on the phase space reconstruction of welding current in short-circuiting transfer arc welding under carbon dioxide, the approximate entropy of welding current and its standard deviation have been calculated and analyzed at different welding speeds and different electrode extensions respectively. The experimental and calculated results show that at a certain arc voltage, wire feeding rate and gas flow rate, welding speed and electrode extension have significant effects not only on the approximate entropy of welding current, but also on the stability of welding process. Further analysis proves that when the welding speed and electrode extension are in an appropriate range respectively, the welding current approximate entropy attains maximum and its standard deviation minimum. Just under such circumstances, the welding process is in the most stable state.     

基于NSGA2与FIuent耦合的分析方法翼型优化

将分析方法引入飞行器气动优化设计,在FIuent软件中进行二次开发,编写了计算熵产率的UDF,来阐述分析方法在气动优化中的作用。以二维翼型为例,通过NURBS曲线进行翼型参数化建模,将NSGA2优化算法与CFD计算耦合起来,计算低雷诺数、5°攻角下的翼型升阻比和熵产率,得到最大升阻比和最小熵产率不断调和的翼型的Pareto解。结果显示,与基准翼型相比,在升阻比提高的条件下,流场熵产率减少,能量效率提高。而且优化翼型的熵产率随着升阻比的增加而增大。     

岸站式河湖水质远程监控柜设计与开发

随着我国工业化进程的加速,河湖水体污染现象日益加剧,已经严重威胁到我国的经济和社会发展。鉴于目前河湖水质监测手段落后和水质评价困难的现状,本文设计了岸站式河湖水质远程监控柜。该监控柜主要安放于河湖岸边,利用水质传感器实时采集河湖的水质信息参数,连同监控柜的状态信息等通过无线网络上传至上位机监控平台,对采集到的水质信息进行分析处理,实现了对河湖水质状况的有效监测,且兼具门禁指纹认证、枯水检测等功能。同时,为了对河湖水质状况做出及时、客观的评价和分级,结合改进层次分析法与熵权法的特点,提出一种水质参数权重确定算法。该算法既兼顾了决策者的偏好,同时又尽量减小了赋权的主观随意性,促使主观与客观评价相统一,且计算较简便、可操作性好。将上述算法嵌入到监控柜上位监控系统中,在唐山青龙河水域进行了系统测试,测试结果表明该监控柜系统运行正常,可以实时获取青龙河的水质信息参数,对河湖水质状况做出及时、客观的评价和分级,实现对河湖水质的综合分析与管理,进而为河湖水环境保护提供一种有效地辅助决策平台。     

基于平均熵的自适应人工蜂群算法

针对基本人工蜂群算法容易陷入局部最优和早熟等问题,提出一种改进的人工蜂群算法(ASABC)。利用平均熵机制初始化种群,增加种群的多样性,避免算法陷入早熟;同时,采用自适应调节邻域搜索步长的策略来提高算法的局部搜索能力,提升算法的计算精度;为了平衡算法的全局搜索能力和局部搜索能力,引入自适应比例选择策略来代替人工蜂群算法的适应度比例选择方法。对8个标准测试函数的仿真实验结果表明,与3种常见的智能优化方法相比,改进的算法具有显著的局部搜索能力和较快的收敛速度。     

熵能量堆煤图像的定位与识别

目的 针对煤矿皮带堆煤传感器受复杂环境干扰导致检测效率较低的问题,提出一种新的基于熵能量(EE)的堆煤图像定位与识别方法。方法 该方法首先在堆煤图像初始化区域内设定采样单元,统计采样单元前景和背景区域的信息熵,计算信息熵分布的极大值所对应的的灰度阶并统计其出现的概率;然后计算采样单元的熵能量和能量频率,归一化频率系数,并利用能量频率的二阶偏导确定堆煤边缘区域,将采样单元移动至堆煤边缘处;最后,根据采样单元的坐标位置计算堆煤高度和半径,进行实时预警。结果 在煤矿井下皮带堆煤视频上进行算法测试,该算法的单帧图像平均处理速度为0.35 s/帧,堆煤事故预警时间间隔最大为0.35 s,平均识别准确率为97.1%,测试结果表明,该方法对堆煤模糊边缘、环境噪音具有抗干扰特点,可以有效克服强噪音、模糊边缘等因素对堆煤识别的影响。结论 该方法对复杂环境中堆煤定位与识别具有很好的实时性和准确性,对提高煤矿安全生产效率具有重要意义。     

装备故障样本优化方法研究

测试性验证装备的故障样本往往相互关联,全部注入费用较高、代价较大.为了降低验证试验费用,采用适当方法对故障样本进行优化分析.为提高故障检测率,提出了一种等价样本的故障样本优化方法.方法在分析故障-测试关联矩阵及其扩展、故障模式功能等价集合和故障模式测试等价集合的基础上,构建了故障样本等价集合,并进行重要度特征分析和最大熵求解,确定了最小的故障样本集合.通过对某型试验台故障样本优化实例分析,并与传统的方法进行了试验结果对比分析,使得故障样本数量、试验费用大为减少,提高了测试性验证的经济性.     

TSS （Tourism Support System） as a New Competitive Tourism Marketing Instrument

The recent trend in tourism marketing is focus on customer relationship management. Tourism industry today is one of the highest revenue generating industry and strategic approach for sustainable development of this industry hosts benefits not only for the tourism related stakeholders but also the community and economy on the whole. Oman is one of the most preferred destinations for the tourists especially after the declaration of Muscat, Arab tourism Capital for 2012. Thus to materialize this honor and position, the scientific study and analysis of tourist behavior will help to predict the future trend of tourism and will give direction for effort investment. This work presents a novel strategy to identify, analyze and highlights the main tourist behavioral factors that could increase tourists＇ loyalty to a specific destination or agency. The analytical TSS （tourism support system） will also classify customers into two categories--first category will be classified as ＂LT （loyal tourists）＂ and second category as ＂NLT （non-loyal tourists）＂. Dataset is collected from a tourism business organization. Twenty-four attributes and 545 instances were collected and were analyzed by algorithms like logistics, forest of random trees, naive Bayes, J48 and Id3. The explanatory variables were defined, and some transformations were done to identify the response variable. Entropy was used and adapted in order to find the response variable from the explanatory variables. The results obtained from this work confirm that the generated rules can be used for future prediction and tourism business can be improved and efforts can be directed in right place for the right consumer resulting in high return on investment.