preisach

粒子群算法是求解函数优化问题的一种新的进化算法，然而它在求解高维函数时容易陷入局部最优，为了克服这个缺点，提出了一种新的粒子群算法，算法对粒子的速度和位置更新公式进行了改进，使粒子在其最有位置的基础上进行位置更新，增强了算法的寻优(PSO for solving function optimization problems is a new evolutionary algorithm, but it is easy to solve high-dimensional function into local optimum, in order to overcome this shortcoming, a new particle swarm algorithm, algorithm on the particle velocity and location of the update formula has been improved, so that particles in the most positions on the basis of location updates, enhancements, optimization algorithm)

electrical power converter

莫尔斯码的幅度调制与解调:信号x(t)含有一段简单的消息，由文件ctftmod.mat定义，由x(t)解出所需信息(The amplitude modulation and demodulation of the Morse code:the signal x (t) contains a simple piece of information, as defined by the file ctftmod.mat information required by the solution x (t))

LMD实验程序，效果不太理想，与大家讨论共享。只能分享出PF1.不能实现PF2,PF3的分离(LMD experimental procedure, the effect is not ideal, discuss and share. Only share the PF1. Separation can not be achieved PF2, PF3 the)

bulit base on ISCAS paper example to reference to paper, we can use the tool to repeat the action of the author(it s an assistance toolbox of the paper to understand what the author say)

说明：  用矩阵实现而微小波变换，解决途中处理中而微信息量过大的问题(Realization with small wave transformation matrix to address the way in which micro-processing problem of information overload)

相关系数计算的模拟,用于衡量两个数列的相关程度(coherence coefficient calculation)

利用两层BP神经网络训练加权系数。隐含层的激活函数取S型传输函数，输出层的激活函数取线性传输函数(Weighting factor of two layers of BP neural network training. The hidden layer activation function to take the S-shaped transfer function, the output layer activation function to take a linear transfer function)

emd分解，并改进后通过镜像拖延来进行端点抑制，端点效应得到改善(The emd decomposition, and the improved endpoints through mirror delay suppressed, endpoint effect improved)

CT滤波反投影重建算法，利用卷积计算，利用矩阵计算积分，比较简单的思想(CT reconstruction)