sparse_decon

在matlab环境下，实现一个雷达仿真程序。(in Matlab environment, the realization of a radar simulation.)

读取segy数据、文件，写道头卷头等等，内置说明书，简单易用(Read the segy data, file, writes head roll and so on, the built-in instructions, easy to use)

matlab音频处理工具箱colea，仅参考及学习。(matlab audio processing toolbox colea, only a reference and learning.)

概率神经网络机械故障诊断matlab程序 通过振动信号预判机械故障(Diesel Engine Fault Diagnosis)

fastica定点独立分量分析方法非线性函数选择tanh(a1y)(fastica fixed-point independent component analysis of nonlinear function selection tanh (a1y))

目的：SAR雷达在斜视角etha下，雷达回波信号仿真 目标位置 X1(0.75PRT*C/2,Ta*V),X2(0.75PRT*C/2,2.3Ta*V) 修改时间：2013年9月21日 11:30

近似熵是一种衡量衡量非线性时间序列复杂性的非线性动力学分析方法，其物理意义 为熵值越大，时间序列的自我相似性越低，序 列越复杂，反之序列自我相似性越高；并且样 本熵对于序列长度依赖弱，仅需较短的数据就 能够得到稳健的熵值，同时其计算不需对数据 粗粒化，具有较好的抗干扰能力 (Approximate entropy is a measure to measure the complex nonlinear dynamics of nonlinear time series analysis method, the physical meaning of entropy value is larger, the lower the self-similarity of the time series, the more complex the sequence, whereas sequence of self-similarity the higher and the sample entropy weak dependence for the length of the sequence, only shorter data can be healthy entropy value, calculated without data coarse-grained, with good anti-jamming capability)

先归一化输入输出参数，用KMens优化RBF神经网络中心值，然后计算隐节点数据中心间的距离（矩阵），得到各隐节点的扩展常数（宽度），接着用最小二乘法得到各隐节点的输出权值，对数据进行反归一化，并绘图得到神经网络输出与测试集图，同时进行性能评价(First, the input and output parameters are normalized, the central value of the RBF neural network is optimized by KMens, then the distance (matrix) between the data centers of the hidden nodes is calculated, and the extended constant (width) of the hidden nodes is obtained. Then the output weights of the hidden nodes are obtained by the least square method, and the data are back normalized, and the neural network is drawn. Output and test set diagrams, and performance evaluation)

说明：  二维码检测 用matlab实现的一个简单程序识别特定二维码(qr code detection, a simple algorithm using matlab)

读mp3文件格式的matlab 源程序(Reading of mp3 file format matlab source)