用于单片机的DSP算法库（查找、排序、复数、向量、矩阵、滤波、统计）

[huangzj121]

用于单片机的DSP算法库（查找、排序、复数、向量、矩阵、滤波、统计）

大多数电子产品里面跑的还是MCU而不是DSP，由于经常要作一些信号处理，所以把一些通用算法做成模块，成本比DSP低了很多，效率上DSP没有乘法硬件也只是提供汇编库，这里所有操作用指针代替。

查找与排序
http://www.math86a.icoc.cc/nd.jsp?id=8&moduleId=3&fromColId=4

复数
http://www.math86a.icoc.cc/nd.jsp?id=9&moduleId=3&fromColId=4

向量与矩阵
http://www.math86a.icoc.cc/nd.jsp?id=10&moduleId=3&fromColId=4

滤波与统计
http://www.math86a.icoc.cc/nd.jsp?id=11&moduleId=3&fromColId=4

/*********************************************************************

++++++++++++++Copyright   Math86A++++++++++++++
Numerical Recipes and Digital Signal Processing

Url:        http://www.math86a.icoc.cc
Tel:        [phone]
Email:      [e-mail]

 

   h_sorts

   This program is include some basic algorithm of find and sort.
   Contain: sequence find, 
            binary find, 
            insert sort, 
            shell sort, 
            select sort, 
            heap sort, 
            bubble sort, 
            quick sort, 
            merge sort,
            a random quick sort, 
            system quick sort.

   Author:    huang zhen jiang
   Date:      2012 / 08 / 10
*********************************************************************/

/*************************************************************

++++++++++++++Copyright   Math86A++++++++++++++
Numerical Recipes and Digital Signal Processing

Url:        http://www.math86a.icoc.cc
Tel:        [phone]
Email:      [e-mail]

 

   h_fcomplex

   This program is include some basic calculate of complex.
   Contain: create one complex, 
            complex addition, 
            complex substruction, 
            complex multiplication, 
            complex division, 
            complex's conjg,
            complex's mod,
            real multiply complex.

   Author:   huang zhen jiang
   Date:     2012 / 08 / 16
*************************************************************/

/*****************************************************************

++++++++++++++Copyright   Math86A++++++++++++++
Numerical Recipes and Digital Signal Processing

Url:        http://www.math86a.icoc.cc
Tel:        [phone]
Email:      [e-mail]

 

   h_fmatrix            
                
   This program is include some basic calculate of real matrix. 
   Contain: creat integrate vector,
            creat float point vector,
            creat double vector,
            print vector,
            delete vector,
            maximum of one vector,
            minimize of one vector,
            copy one vector,
            vector's addition,
            vector's substruction,
            vector's multiplication,
            vector multiply a real,
            create zero matrix,
            create one matrix,
            create identity matrix,
            print matrix,
            delete matrix,
            copy one matrix,
            matrix's transform,
            matrix's addition,
            matrix's substruction,
            matrix's multiplication,
            matrix's inverse,
            matrix's division,
            matrix multiply a real.
               
   Author:   huang zhen jiang     
   Date:     2012 / 09 / 16     
*****************************************************************/

/******************************************************************************************

++++++++++++++Copyright   Math86A++++++++++++++
Numerical Recipes and Digital Signal Processing

Url:        http://www.math86a.icoc.cc
Tel:        [phone]
Email:      [e-mail]

 

   h_filter

   This program is include some basic calculate and algorithm of digital signal process.
   Contain: fixed point convolution,
            float point convolution,
            fixed point filter difference function, 
            float point filter difference function, 
            calculate data's maximum,
            calculate data's minimize,
            calculate data's average,
            calculate data's integral,
            calculate data's summation,
            calculate 1-dimension discrete-fourier-transform,
            calculate 1-dimension inverse discrete-fourier-transform,
            calculate 1-dimension fast-fourier-transform,
            calculate 1-dimension inverse fast-fourier-transform,
            calculate cross correlate,
            calculate variance,
            calculate correlate coefficient,
            calculate power spectrum,
            LMS adaptive filter,
            RLS adaptive filter,
            kalman adaptive filter,
            kalman estimate.

   Author:   huang zhen jiang
   Date:     2012 / 08 / 19
******************************************************************************************/

[huangzj121]

线性查找、折半查找、插入排序、希尔排序、选择排序、堆排序、冒泡排序、快速排序、归并排序、一个随机化实现的快速排序、系统快速排序。

算FFT和PSD功率谱用得着。生成复数，复数加、减、乘、除、共轭、求模、乘实数。

自适应LMS、RLS、kalman都是向量和矩阵运算。生成向量，向量极值，向量复制，向量加、减、乘、点乘，生成单位阵，复制矩阵，矩阵转置，矩阵加、减、乘、除、求逆（高斯约尔当法，很实用的东西）、点乘。

FIR卷积和，IIR差分方程，数据积分，DFT，FFT（由于1024点以下基2、基4以及分裂基方法数量级相差不是很多，而且基2方法容易程序实现，所以一般都采用基2法），互相关，方差，相关系数，功率谱，LMS自适应滤波，RLS自适应滤波，卡尔曼滤波，卡尔曼估计。

[huangzj121]

很多工程师只知道用DSP，用DSP，

真正懂DSP吗？

我就奇怪了，DSP只是用来中断、指令、时钟这种东西吗？

只是用一些大公司封装好的CHIP来混弄领导、糊弄顾客、也混弄了自己大好的青春年华，

我看一般搞DSP的没几个愿意搞切比雪夫逼近，

小波函数是个啥东西，我也没弄懂，

浮躁的国人

[woshi_ziyu]

如果考虑使用算法 最好是使用DSP 不采取在MCU上实现

[huangzj121]

不光ST，PIC，LPC都可以用

[gaopeng090610]

学习下，谢谢楼主·····