Firdesi

基于小波包分解的语音去噪，根据熵谱概率密度函数估计阈值，去除实际环境噪声。(Based on wavelet packet decomposition of the voice de-noising, in accordance with spectral entropy probability density function estimated threshold, the actual removal of environmental noise.)

Dual kalman filter for battery charge estimation

基于matlab/Simulink的动态仿真与控制的学习资料，有多章的相关学习内容。(Dynamic simulation and control matlab/Simulink learning materials, a number of chapters related learning content.)

利用无迹卡尔曼滤波的平方根形式做的一个简单例子(A sample case of using SRUKF to find the true value)

白噪声中单个正弦信号的频率检测与估计 产生，其中是一高斯白噪声，其均值为0，方差为1。用MUSIC方法估计观测数据中正弦波的频率，并给出白噪声方差()un2uσ 与复正弦波的振幅A的估计值。(A single frequency detection and estimation of the sinusoidal signal in white noise, which is a Gaussian white noise with mean 0 and variance 1. MUSIC method to estimate the frequency of sine wave in the observation data, and gives the white noise variance () estimated value of the the un2uσ complex sine wave amplitude A.)

Applying postureidentifierindesigninganadaptivenonlinearpredictive controller fornonholonomicmobilerobot

自己实现的，模糊C均值聚类的代码，在K-mean上添加了隶属度矩阵。注释详细，方便理解算法步骤(my implementation, Fuzzy C-Means clustering code in the K-mean adding a membership matrix. Notes detailed, easy to understand algorithm steps)

说明：  ：将信息论中熵的概念应用到特征选择中，定义了两种信息测度评价特征——误差熵和混叠熵，然后阐述了两种定义的不 用物理意义，分析了计算熵中最关键的区间划分问题，并提出一种较好的区间划分方法。(: The concept of entropy in information theory applied to feature selection, the definition of information measure evaluation of two features- error entropy and mixing entropy, and then describes the two definitions do not have the physical meaning of the calculation of entropy in the range of the most critical partitioning problem and propose a better method of interval division.)

NNIAToolbox es un algoritmo para evaluar problemas multi objetivo obteniendo frentes de pareto

人工神经网络有着广泛的应用，这段代码给出了NN在人脸检测中的具体应用实例。(face detection )