200503-95

一种改进的LMS算法及其在噪声对消中的应用。在分析传统定步长LMS算法和变步长LMS算法的基础上，提出了一种改进的变步长LMS算法。新算法利用瞬时误差绝对值三次方的指数形式和遗忘因子同时调整步长，更好的解决了收敛速度和稳态误差的矛盾(An improved LMS algorithm and its application in noise cancellation. In the analysis of the traditional fixed step and variable step size LMS algorithm based on LMS algorithm, a modified variable step size LMS algorithm. Absolute value of the new algorithm uses the instantaneous error and the form of cubic exponential forgetting factor while adjusting step, a better solution to the convergence speed and steady-state error of the contradictions)

这是我强烈推荐的一本书，它是基于开源软件做的类似于matlab的一个工具，功能非常的强大，建议在校学生好好学习这门课，对于以后的应用会有很大的帮助(python)

径向基网络人工神经网络程序（聚类算法）,应用MATLAB 软件开发。(RBF-ANN)

DMC控制 当系统的参数知道很少时可以利用这种方法进行运算 从而很好的控制(DMC control parameters when the system knows very little use of this method can well control operations to)

To simulate RS and physical layer cross-layer design mechanism in a fading channel

混沌动力学的平面模型，和对动力学的模型研究以其用户的输入(chaoic reserch ，the model ,and teh input model .)

file is containing the exponential distribution and random numbers

根据遗传算法和BP神经网络理论，在MATLAB软件中编程实现基于遗传算法优化的BP神经网络非线性系统拟合算法。(According to the theory of genetic algorithm and BP neural network, in the MATLAB software programming to realize BP neural network nonlinear systems based on genetic algorithm optimization fitting method.)

利用Djikstra算法计算最短路，输入的参数为一个矩阵，可以是无向图或者有向图(calculate the shortest distance)

经验模态分解（EMD）方法（平稳化过程）是一种全新的处理非平稳数据序列的方法，其基本思想是：假如一个原始数据序列X(t)的极大值或极小值数目比上跨零点（或下跨零点）的数目多两个（或两个以上），则该数据序列就要进行平稳化处理(Empirical Mode Decomposition (EMD) method (a smooth process) is a new deal with non-stationary data sequence method, the basic idea is: If a raw data sequence X (t) the maximum or minimum number of than the cross-zero (or under the cross-zero) number more than two (or more), then the data is necessary to carry out a smooth sequence of treatment)