allmargin

基于PD增益自适应调节的模型参考自适应控制(based on PD Gain adaptive regulation of MRAC)

能对压缩感知中的稀疏度未知的信号精确的重建，重建误差小，经本人调试，完全正确(Capable of compressed sensing signals unknown sparsity accurate reconstruction, reconstruction error is small, after I debug, completely correct)

关于狭缝衍射的夫琅和费远场模拟的例子，是高等物理光学模拟的入门例子。(About slit Fraunhofer diffraction far-field simulation example is the example of the entry of higher physical optics simulation.)

a matlab code to reduce the dimenision of data. using probability principle analysis

说明：  根据脉冲序列，使用s域z域之间的转换，求系统的脉冲响应传递函数(According to pulse sequence, the use of s-domain conversion between z-domain, and the system impulse response transfer function)

使用相关函数来检测信号序列中隐含的周期性(Using the relevant function to detect the implicit periodic of signal sequence.)

Machine learning and pattern recognition.

利用牛顿法，加速牛顿法，同伦变换求解非线性方程组的解(Newton method to accelerate the Newton method, homotopy transform Solving the Nonlinear Equations)

通过 matlab 编程实现捷联惯导系统初始对准、惯导解算的全过程。 1）初始自对准：利用前10分钟实验车静止的数据进行惯导系统的自对准。采用前2分钟数据进行解析粗对准，后8分钟数据用五状态Kalman滤波器实现精对准；为了验证精对准性能，可在粗对准结果上把航向角人为加一定幅度（如10度）的误差，观察精对准的收敛性能。 2）纯惯导解算：对准结束后转入纯惯性导航解算，要求进行双子样的圆锥和划摇补偿，即每20ms完成一次导航解算，输出完整的导航结果包括三个姿态、两个水平速度和两个水平位置（天向通道发散不必输出），并给出姿态、速度、位置随时间的变化曲线以及位置误差曲线。(The whole process of initial alignment and inertial navigation solution of strapdown inertial navigation system is realized by matlab programming.)

PF,EPF,UPF的对比仿真代码，可以很直观的看出各种算法之间的差别及各自的优缺点。可直接运行。(PF, EPF, UPF contrast simulation code, it is intuitive to see the differences between the various algorithms and their respective advantages and disadvantages. Can be directly run.)