下载的几个C语言版的FFT程序仿真移植成功，特分享-第9页回答

下载的几个C语言版的FFT程序仿真移植成功，特分享

2020-01-13 18:37发布

该问题目前已经被作者或者管理员关闭, 无法添加新回复

86条回答

millwood0

1楼-- · 2020-01-19 14:06

the code compiled and tested on turbo c, but should work on any other c compilers as well.

millwood0

2楼-- · 2020-01-19 17:13

the problem with floating point math on a 8-bit mcu is its size / speed. so I adapted a fixed point fft routine.

it completes 32-point fft in 20k cycles and 128-point fft in 107k cycles (on a 24f chip, no optimization and no hardware acceleration) - a ~10x improvement over the floating point math version.

again, the package consists of three files:

1) fft_fp.h: the header file. you need to configure again the buffer size + data type.
2) fft_fp.c: the source file. you need to include it in your project.
3) main.c: the test program.

the calling convention is (almost) identical to that of the floating math version.

millwood0

3楼-- · 2020-01-19 18:17

here is the package:

点击此处下载 ourdev_716318BSJNK0.rar(文件大小:9K) (原文件名:24f fft_fp.rar)

millwood0

4楼-- · 2020-01-19 23:00

I am on a roll!

here is a kalman filter, for 1d variables (aka only one state variable). It is useful for filtering, and or tracking.

the package consists of two files:

1) kalman.h: the header file declaring the functions;
2) kalman.c: the source file laying out the actual code / execution.

in additional, I put together a main.c calling both functions to show how it could be used.

here is the kalman.h:

#ifndef KALMAN_H_INCLUDED
#define KALMAN_H_INCLUDED

//kalman filtering for 1d variables
//source: http://interactive-matter.eu/2009/12/filtering-sensor-data-with-a-kalman-filter/

typedef struct {
  double q; //process noise covariance
  double r; //measurement noise covariance
  double x; //value
  double p; //estimation error covariance
  double k; //kalman gain
} kalman_state;

//initialize the kalman filter
kalman_state kalman_init(double q, double r, double p, double intial_value);

void kalman_update(kalman_state* state, double measurement);

#endif // KALMAN_H_INCLUDED

millwood0

5楼-- · 2020-01-20 04:40

精彩回答 2 元偷偷看……

millwood0

6楼-- · 2020-01-20 07:53

here is main.c: it uses a function to generate sampled data.

#include <stdio.h>
#include <stdlib.h>
#include "kalman.h" //we use kalman filtering

//generate simulated data, with a noise term
double x_actual(void) {
static double xt_1=-10, xt_2=20;
double xt = 1.2*xt_1 - 1.0*xt_2 + 00.0 * rand()/RAND_MAX;

xt_2 = xt_1;
xt_1 = xt;
return xt;
}

int main()
{
FILE *fptr; //file ptr
int i;
kalman_state k_state;
double xt;

//open the file
fptr = fopen("c:kalman.dat", "w");
//fprintf(fptr, "abcdef ");

k_state = kalman_init(1, 1, 1, 1); //initialize the filter

printf("Hello world! ");

for (i=0; i<200; i++) {
xt = x_actual(); //generate xt's measurement
kalman_update(&k_state, xt); //estimate xt
fprintf(stdout, "i=%4d x_actual = %10.4f x_est = %10.4f ", i, xt, k_state.x);
}

fclose(fptr); //close the file
return 0;
}

首页上一页 7 8 9 10 11 12 13 14 下一页最后一页

下载的几个C语言版的FFT程序仿真移植成功，特分享

一周热门更多>

相关问题

相关文章

下载的几个C语言版的FFT程序仿真移植成功，特分享

打开微信“扫一扫”，打开网页后点击屏幕右上角分享按钮

一周热门 更多>

相关问题

相关文章

采纳回答

编辑标签

举报内容

检举类型

检举原因

检举说明(必填)

一周热门更多>