极小值

Because of appearing empty aggregates in the time zone (the time zone do not include the information), we only establish the indeterminate grey differential equation according to the approximate condition of the differential equation.

利用灰色微分方程中的最小二乘估计参数：进而得到灰色神经网络初始的权值及阈值，这样使初始的权值和阈值更接近准确值，克服了以往BP网络中初始权值的随机性和不确定性，减少了网络的训练次数，提高了网络收敛速度和输出的精度，为避免网络训练陷入局部极小值也起到了重要的作用。

Cold and hot appetizers 1、steamed fresh scallop seasonal price 2、deep fried lobster in salt and chilli seasonal price 3、braised lobster with ginger and spring onions seasonal price 4、crispy sqring rolls 5、sesame prawn on toast 6、grilled prawn in salt and chilli 7、paper prawn on toast 8、king prawn cutlet in breadcrumbs 9、prawn cocktail 10 sweet and sour won ton 11 deep fried minced squid with thai sauce 12 deep fried squid in salt and chilli 13 deep fried squid 14 barbecue spare ribs 15 mini spare ribs in o.k sauce 16 mini spare ribs in honey sauce 17 mini spare ribs in salt and chilli 18 capital spare ribs 19 satay chicken 20 szechuan style lettuce wraps 21 pang pang chicken 22 shredded chicken in yellow bean sauce 23 paper wrap chicken 24 crispy seaweed 25 crispy lamb wraps with lettuce 26 griddled fried pork dumblings 27 steamed pork dumplings 28 stir fried mussels in black bean sauce 29 special crispy platter(minimum for 2 people)(spring rolls satay chicken crespy seaweed.barbecue spare ribs and sesame prawn on toast) 30 seafood platter (minimum for 2 people) sesame prawn on toast deep fried seallops.

冷和热的开胃菜 1、steamed新鲜的扇贝季节性价格 2、油炸了在盐和辣椒季节性价格的龙虾与姜和春天葱季节性价格的3、braised龙虾4、crispy sqring的卷在多士的5、sesame大虾在盐和辣椒的6、grilled大虾在多士的7、paper大虾 8、king在面包渣的大虾炸肉排 9、prawn鸡尾酒 10糖醋被赢取的吨 11被油炸的剁碎的乌贼用泰国调味汁 12在盐和辣椒的被油炸的乌贼 13被油炸的乌贼 14烤肉排骨 15在好的调味汁的微型排骨 16在蜂蜜调味汁的微型排骨 17在盐和辣椒的微型排骨 18资本排骨 19 satay鸡 20 szechuan样式莴苣套 21剧痛剧痛鸡 22在黄豆调味汁的切细的鸡 23纸套鸡 24酥脆海草 25酥脆羊羔套用莴苣 26烤了油煎的猪肉dumblings 27蒸的猪肉饺子 28混乱油煎了在黑豆调味汁的淡菜 29特别酥脆盛肉盘(2个人的极小值)(春卷satay鸡在多士的crespy seaweed.barbecue排骨和芝麻大虾) 30海鲜盛肉盘(2个人的极小值)在多士被油炸的seallops的芝麻大虾。

The numerical results indicated that the mixing time decreases about linearly with increasing current intensity. The mixing time decreases first then increases with increasing current frequency and comes up to its minimum value at 30Hz. Applying the traveling magnetic field to the uptake or downtake when the exciting current intensity is 200 A at 10 Hz, the mixing time can be shortened by 9%~17%. With the operating conditions kept unchanged, shorter mixing time is available if applying the traveling magnetic field to the uptake instead of downtake, and the time can be shortened by 18%~26% if applying the magnetic field to both the uptake and downtake. Moreover, the mixing time decreases with increasing argon blowing flowrate, and the time comes down to its minimum value if the argon blowing flowrate is 1 600 NL/min. After the moment, the mixing time increases with argon blowing flowrate.

计算结果表明：混匀时间随电流强度的增大而减小，并且近似成线性关系；混匀时间随电流频率的增加先减小后增大，并且在30Hz 时达到极小值；在励磁电流强度为200A ，频率为10Hz 的条件下，在上升管或下降管施加行波磁场混匀时间可缩短9%~17%；在相同操作条件下，在上升管处施加行波磁场混匀时间小于在下降管施加行波磁场的混匀时间；同时在上升管和下降管施加行波磁场，混匀时间可缩短18%~26%；混匀时间随吹氩量的增大而减小，在吹氩量为1600NL/min 时混匀时间达到极小值，吹氩量继续增大时，混匀时间反而增大。

Aimed at the characteristic of multiple types of faults possibly happened in nuclear power plant, large scale of training sample, and requirement of quick and accurate diagnosing, after the analysis of large sample number generated by large fault type exceeds limitation of 64K paragraph in DOS and large number of local minmum in error surface, measurements of same format of data file, Win95/NT operating system platform, resetting weight learning rate, dynamic training set in quick learning algorithm and improvement on quick learning algorithm using homotopy method which can avoid local minmum points in error surface have been adopted to ensure quickly and effective process of the course of neural network's training and testing.

针对核动力装置可能发生的故障种类多，训练样本规模大，故障诊断需要快速准确的特点等，分析了样本多和故障种类多产生的训练样本量超出DOS段大小和网络误差曲面上局部极小值多的情况，采取了使用相同格式数据文件、Win95/NT操作系统平台，对快速学习算法采用重置神经网络权值学习率和动态训练集、并采用能够有效克服网络误差曲面上局部极小点的同伦方法对学习算法进行改进等措施保证神经网络训练测试过程的快速、有效地进行；针对装置发生的故障须快速、准确诊断的需要，分析了故障的产生对装置参数变化的影响及操纵员对故障诊断的基础，在核动力装置发生故障时参数曲线的变化量与正常运行时参数曲线的变化量存在明显差异的基础上，提出采用参数曲线的变化量作为神经网络的输入，并围绕参数变化量的方法采用二次曲线拟合滤波求变化量和阈值技术来保证神经网络得到精确的装置参数变化量，从而得到准确的诊断结果。

It is difficult to obtain the accurate segmentation threshold of the image. This paper uses the multinomial to approximate the histogram envelope curve by different phase. So the question of how to obtain the accurate segmentation threshold of the image can convert into the solution of the minimum of multinomial, Using this method it can be solved the question how to segment the image more accurately.

对于在图像分割时难于得到图象分割的准确阈值问题，提出利用多项式对直方图包络线进行分段拟合，把图像分割阈值问题转化为求多项式极小值的问题，从而求出用于分割图像不同灰度层次的最优阈值的方法。

In summary, the drive current by Alfvén waves can be enhanced by plasma rotation The negative magnetic shear is favorable for the efficiency of the current drive by Alfvén waves Moreover, in the reversed shear configuration the current drive due to Alfvén waves in an rotation plasma can obviously change the plasma current profile The Alfvén wave current drive may be useful in controlling the current profile in order to suppress magnetohydrodynamic instabilities.

另外，我们还发现负剪切区域愈大，同时安全因子极小值q〓愈小，对阿尔芬波电流驱动愈好。总之，等离子体极向旋转能够增强阿尔芬波电流驱动，负磁剪切对阿尔芬波电流驱动是有利的。另外，反磁剪切位形下旋转等离子体的阿尔芬波电流驱动可以明显改善等离子体电流剖面。

The results showed that there were three minimum values in the second derivative spectrums of normal person, in contrast there were four minimum values in most diabetic second derivative spectrums, the diabetic urine had one minimum value more than normal urine during the range from 550 nm to 570 nm.

发现健康人的二阶导数光谱都有三个极小值，与之形成对比的是绝大多数糖尿病患者的二阶导数光谱却有四个极小值，糖尿病患者的尿样在550 nm～570 nm区间范围内比健康人多出一个极小值。

A Study on Fraunhofer Double-Slit Diffraction by Applying Simulation Technology;2. Based on Kirchhoffs theory,the approximate intensity formulas of double-slit diffraction of Gaussian Beam is derived under paraxial approximation with power series expansion methods.

依据基尔霍夫衍射理论，利用幂级数展开的方法推导出旁轴条件下高斯光束双缝衍射的近似光强计算公式，并通过数值计算对双缝衍射的基本特性进行了讨论，得出了与其他文献一致的结论，得出了高斯光束下的双缝衍射极小值点比平行光束双缝衍射的极小值点向外移，观察屏中心光能变小；光强分布可由激光斑在衍射屏处的参数调制的结论。

In the studying of image segmentation, six crucial technologies havebeen developed successfully. 1 In one dimension histogram of original DRimage of luggage, there remain maximum-value wave crest havingapproximately stationary position, and the first minimum to the left of thewave crest performs excellently to be used to get rid of background of DRimage. 2 Open operation, one type of math morphology, is the right idea toerase luggage disturbance in DR image and at the same time, can be retaininteresting region appropriately. 3 After open operation, there are only a few minimums on image histogram, one of which can be to segment image furtherand achieves satisfying result. 4 Method of histogram analyzing and that oferosion performs well in the re-segmentation of conglutinative region in DRimage. 5 Method of image distance switch can ensure Luggage CT sectionautomatically obtained to have rich suspicious region information. 6 Todecompose structure element of math morphology is a way to save a great dealrunning time for automatically discerning interesting regions algorithm,apparently improving implementation efficiency.

在对图像可疑区域分割研究中，主要开发成功以下几项关键技术：1发现原始箱包DR图像的一维直方图中，均具有位置相对稳定的最大值波峰，以波峰左边第一个极小值为分割阈值，去除DR图像背景，效果非常好。2运用数学形态学中的开运算能够非常有效地去除箱子的干扰，同时能够比较好地保留可疑区域。3发现开运算后的DR图像直方图只有若干个极小值点的特点，其中用最大值波峰左边的第二个极小值点进一步分割DR图像，能够更有效地提取箱包中的可疑区域。4运用直方图分析法和腐蚀法，对粘连区域实行再分割，效果较好。5运用距离变换法，能确保自动设定的箱包CT截面包含丰富的可疑区域信息。6运用形态学结构元素分解技术，大幅度减少可疑区域的自动判别算法运行时间，提高算法软件执行效率。

The shortages of typical BP algorithm are that its convergence speed is slow and it possibly mires to local minimum. To solve the former shortage, BP algorithm was modified from several aspects, such as adjusting learning rate adaptively, adding momentum factor, preprocessing training data and modifying network structure, etc.. For the latter shortage, factor γ was introduced to avoid most of local minimum and accelerate the convergence speed.

为解决现有BP算法收敛速度慢的问题，对BP算法进行了改进，采用了动态改变步长、加入动量因子、对训练样本进行预处理以及调整网络结构等措施；针对经典BP算法可能收敛到局部极小点的问题，加入γ因子，可以避免大部分的局部极小值，从而使网络的收敛速度加快。

As she looked at Warrington's manly face, and dark, melancholy eyes, she had settled in her mind that he must have been the victim of an unhappy attachment.

每逢看到沃林顿那刚毅的脸，那乌黑、忧郁的眼睛，她便会相信，他一定作过不幸的爱情的受害者。

Maybe they'll disappear into a pothole.

也许他们将在壶穴里消失