查询词典 algorithm

Substituting other faster algorithm of NN for the BP algorithm in it, we get a new non-line homotopic synthetic learning algorithm, which increases the speed of convergence and improves the ability of avoiding local minimum.

在此基础上，进一步提出用其它改进的神经网络快速算法代替其中的BP算法，从而形成新的非线性同伦综合学习算法，即保持了快速收敛的特点，又较大地提高其避免陷入局部极小的能力，收到很好的训练效果。

Then, it introduces an adaptive blind watermark algorithm based on odd-even judgement. This algorithm uses the neighboring symbols mean value and the odd-even judgement to embed the watermark into the 2-level subband of the carrier image. The results of the test demonstrate that this algorithm has a good imperceptivity and robustness.

然后，在此基础上介绍了一种基于奇偶判决的图像自适应盲水印算法，该算法利用图像小波系数的相邻特征平均值和奇偶判决法在图像小波域的二阶子带上嵌入水印，试验结果表明该方法具有不错的不可见性和鲁棒性。

To solve the problem, a pre-classification algorithm of measured data is introduced and a new algorithm is developed to divide unmeasured data into determinable and indeterminable data, using matrix projection algorithm.

采用已测数据预分类的方法，对其进行了修正。在此基础上，将矩阵投影算法引入到了未测数据分类中，提出了基于矩阵投影算法的未测数据分类算法。

Two algorithms are put forward to solve the problem of on-board engine torque estimation, which is the key technology of Coordinated Control Algorithm. One is called the algorithm of engine torque estimation based on mean-value model and the other is called the algorithm of engine torque estimation based on instantaneous crankshaft speed.

在仿真研究的基础上，针对如何实现动态协调控制算法中的关键技术——发动机转矩在线估计提出了两种算法：基于发动机平均值模型的转矩估计算法和基于发动机曲轴瞬时转速的转矩估计算法。

Based on the principle of Otsu method with maximum variance between threshold algorithm of image segmentation, an improved method derived from Otsu algorithm is put forward, which combines interclass distance with intraclass distance, a partial recursive algorithm is used to search optimum threshold.

基于最大类间方差阈值图像分割算法的基本原理，然后结合目标与背景两类之间间距和类内距离对图像分割效果的影响，提出了一种改进的最大类间方差法，运用递归思想局部搜索图像的最佳阈值。

Based on improved genetic algorithm and primal-dual interior point algorithm, a hybrid reactive power optimization algorithm is proposed.

基于改进遗传算法和原对偶内点法提出一种求解无功优化问题的混合算法。

In order to improve our CNC system to a higher level, an interpolation problem of circular in an arbitrary plane was also studied in the paper, where three kinds of interpolation algorithm were given, their feasibility were analysed and compared, and the best algorithm was obtained; In the end of this paper, a new concept of zero—following error position control was advanced, whose control principle and algorithm wre derived.

为使数控系统向更高档次发展，本文还研究了空间圆弧插补问题，给出了三种插补算法，通过分析比较，得到了最佳算法，并用汇编语言进行了可行性验证；在本文的最后，提出了零跟随误差位置控制的概念，并推导出了控制原理和算法。

In the comparison experiments with genetic algorithm,immune genetic algorithm,basic particle swarm optimization algorithm and so on,IPSO performs well with best solution and fastest convergence speed.

在与遗传算法、免疫遗传算法、基本微粒群等算法的仿真比较试验中，该算法不仅搜索到了最好的近优解，而且收敛速度最快。

The main research methods are:(1)Theory analysis:from the 2D plane geometry analysis, dots continuity characteristics are found. The first time irrational fraction algorithm is used in FM program.(2)Experiments reaearch:experiments of condition halftone function are designed under the cross-experiment principles. The first time variable information printer and C# language are used in experiment.(3)Data processing:new type digital density apparatus is used to measure data. Compare is used in processing data and error cause are explained. The main works as follows:The dot area percentage has been gained and new condition halftone function has been set up.(2)After being corrected, speed of irrational fraction algorithm has been compared with multiplicative congruence algorithm's.

本文的主要研究方法有：(1)理论分析：运用数字加网原理，从二维平面几何分析入手，借助实验验证研究网点连续特性；首次系统地把无理数取尾法随机数生成器引入到调频加网系统中，根据实验修正生成器的模型，运用统计假设检验理论，对生成的随机数序列进行统计假设检验；(2)实验研究：根据正交实验法则设计实验，首次采用可变信息印刷机(HP Laser PS印刷机)输出实验样张，克服了用传统胶印机作实验影响因素多的缺陷；采用新型数字式密度仪测量密度值，提高了数据精度；采用最新编程语言编写调频加网程序；(3)数据处理与分析：利用平均值比较法处理数据，对误差产生的原因进行讨论。

One algorithm,isoperimetric algorithm based on graph theory is applied and researched on image segmentation.Firstly,isoperimetric algorithm is applied in still image segmentation,and the result is compared with two methods based on edge detection and region detection.

1引言图像分割是图像处理领域和计算机视觉中的一个基本而关键的问题，其目的就是将人们感兴趣的目标从图像背景中提取出来，为后续的分类、跟踪、识别、处理等提供基础。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。