值

A modified aerial image was obtained by convolving the aerial image with a Gaussian filter to smear the image in a manner analogous to resistive diffusion, and then applying a variable threshold resist model on the modified image to predict the CD variations.

用高斯滤波器与空间影像进行卷积，得到改进的空间影像来模拟光刻胶的扩散，然后使用可变阈值光刻胶模型，以实际工艺数据拟合该模型参数，再把改进空间影像的相关信息作为输入，并根据可变阈值光刻胶模型所确定的光强阈值来预测CD变化，从而使光刻仿真的结果更精确地附合实际测量数据。

Secondly, in LFRR system flows which weights are the same and larger than 1 are aggregated into a virtual flow. LFRR deals with virtual flows instead of actual flows. This deceases the flow number that scheduling algorithm deals with.

2LFRR算法中采用了等权值流合并的技术，把权值大于1且权值相等的流合并成一个虚流，以虚流为处理对象，使算法需要处理的对象数目大为减小。

As the same section is concerned, near the bottom the maximum value of frequency appear within 1mm of bubble size, towards the water face this value tend to decrease, yet the corresponding size turns to increase; in the same time, the frequency is concentrated near the bottom and deconcentrated near the water face; under the experiment condition, the maximum frequency of the total section approximately appear at 0.5mm size; mean bubble size access 0 at bottom and increase acceleratively towards water face where tend to be infinite; total number increase during developing of aeration, and become steady when reach wholly developed area; on the same section, the total number in unit time tend to increase at first and decrease then, along the water depth.

对于同一断面，靠近槽底的测点气泡个数分布最大值出现在气泡尺寸小于1mm的范围内，朝水面方向各测点气泡个数分布最大值呈减小趋势，而其对应尺寸则呈现变大趋势；在靠近槽底，个数分布表现得比较集中，而在水面附近气泡个数分布曲线表现得低而平缓；实验条件下，整个断面上的气泡个数分布的最大值大体上出现在尺寸为0.5mm的地方；气泡平均尺寸在槽底处接近于0而朝水面方向迅速增大，且越接近掺气水流的水面，增加速度越快，最终趋于无穷大；气泡总数在掺气发展区内是沿程增加的，当掺气水流发展到一定程度后气泡总数就基本保持不变了；在同一断面上，各测点单位时间内的气泡总数随高度呈现出先增加后减小的趋势。

Baidu not only to rich guest source of such high grade UGC undertakes demoting, and the side is being ranked in the different source of same article also is to side with those big stations normally, same an article, inevitable platoon is in the website with authority high cost before the website with authority low cost, although authority is worth the article of low website,just be to be achieved formerly.

百度不但对博客这样的优质UGC源进行降级，而且在同一文章的不同源排名方面通常也是向着那些大站的，同一篇文章，权值高的网站必然排在权值低的网站前面，即使权值低的网站的文章才是原创。

Dichotomized contingency table analysis showed that OR for those positive both for anti-HCV and HBsAg was 61.37,significantly higher than that for single positive of either one,with an additive effect.

叉生分析结果表明，抗-HCV和HBsAg双阳性的OR值为61.37，较两指标单一阳性的OR值明显为高，亦高于两者单独阳性的OR值之和。

The effect of concentration, temperature and different preparation methods on the emulsifying properly has been studied.

本论文研究了不同反应条件下得到的产品的乳化力，初步探讨了反应条件、羟值、皂化值、酸值和产品内部结构与产品乳化力之间的关系。

When water is added in increasingamounts to the normal soil,there is no increase in pH values of the suc-cessive filtrates and the pH of the distilled water is easily obtained,whilstthe pH value of the alkali soils falls very gradually to a ...

水土比例增大时，普通土壤各次滤液的pH 值并不增大，且易于达到蒸馏水的pH 值，反之，碱土的pH 值很慢的降低至pH 7.5。

According to the actual circumstance of performance evaluation for mechanical products, the general neural network is reformed. Therefore, we construct the fuzzy neural network. This paper expresses membership function in the network by trapezium formulae .The indicators are fuzzed using the membership function .

网络中采用梯形函数的四元组形式构造模糊隶属函数，并用隶属函数的端点值和中间值作为样本，用改进的反向传播算法训练网络，得到与待评价产品情况相符合的权重值。

Firstly-the values of voltage on various test points under normal.states and fault states are measured;and the deviation of voltage is fuzzed for deriving the membership degree of the deviation under different fault states.Then.with the membership degree of each test points and their relevant fault categories 88 the sample for training neural network to establish non.linear function from fault feature to fault categories.

首先测得模拟电路在故障状态和正常状态时各测试节点的电压值，并对其电压偏差值模糊化，从而求得不同故障状态下各测试节点的电压偏差值所对应的隶属度；其次，以各测试节点的隶属度和所对应的故障类型作为训练神经网络的样本，来建立从故障特征到故障类型的非线性函数；最后。

Two algorithms breaks through the shortcoming of weighting value immobilizes in the traditional average filter algorithm, and it can characteristic adjust the weights from of data in the filter window, thus makes the weights from noise point is very small contributing to the output date.

两种算法突破了传统的均值算法中加权值固定不变的缺点，可以根据窗口内数据的特点来自适应的调整赋予每个采样点的权值，从而使受噪声污染的信号对输出的权值贡献非常小，而且通过对滤波因子的调整，使得对于脉冲噪声和高斯噪声均有良好的抑制作用。

But we don't care about Battlegrounds.

但我们并不在乎沙场中的显露。

Ah! don't mention it, the butcher's shop is a horror.

啊！不用提了。提到肉，真是糟透了。