最近为了项目，同事让我帮他做一个硬件版的kalman滤波器，实现对设备的kalman滤波，以验证他的理论算法。
犹豫了好几天，用dsp吧，我的kalman滤波算法比较简单，有点大材小用。刚好手里有一块arm调试版，也装了wince系统，就准备在.net环境下编一个kalman滤波器。
虽说学的是导航专业，对kalman滤波应该比较熟悉，可是当时学的就不好，所有学的东西都还给导师了。（导师您不会看到这篇文章吧！看来不能放在首页上！）
于是，只能在网上找一些相关资料。感觉现在变的懒了，总是喜欢在别人的代码上改来改去，不愿意去思考了。算了，反正就这么一次。罪过罪过！
国内的资料对于matlab的算法比较多了，网上随便down，但是基于C#的比较少，还好我的搜索能力不是很差，总算让我找到了相关资料。
引用博客的地址是：http://blog.csdn.net/csdnbao/archive/2009/09/24/4590519.aspx
文章把整个算法都写出来了，我也一起贴出来吧！
using System;
using System.Collections.Generic;
using System.Text;
namespace SimTransfer
{
public class KalmanFacade
{
#region inner class
class KalmanFilter
{
int MP; /* number of measurement vector dimensions */
int DP; /* number of state vector dimensions */
int CP; /* number of control vector dimensions */
public Matrix state_pre; /* predicted state (x'(k)):
x(k)=A*x(k-1)+B*u(k) */
public Matrix state_post; /* corrected state (x(k)):
x(k)=x'(k)+K(k)*(z(k)-H*x'(k)) */
public Matrix transition_matrix; /* state transition matrix (A) */
public Matrix control_matrix; /* control matrix (B)
(it is not used if there is no control)*/
public Matrix measurement_matrix; /* measurement matrix (H) */
public Matrix process_noise_cov; /* process noise covariance matrix (Q) */
public Matrix measurement_noise_cov; /* measurement noise covariance matrix (R) */
public Matrix error_cov_pre; /* priori error estimate covariance matrix (P'(k)):
P'(k)=A*P(k-1)*At + Q)*/
Matrix gain; /* Kalman gain matrix (K(k)):
K(k)=P'(k)*Ht*inv(H*P'(k)*Ht+R)*/
Matrix error_cov_post; /* posteriori error estimate covariance matrix (P(k)):
P(k)=(I-K(k)*H)*P'(k) */
Matrix temp1; /* temporary matrices */
Matrix temp2;
Matrix temp3;
Matrix temp4;
Matrix temp5;
public KalmanFilter()
{
MP = 1;
DP = 2;
CP = 0;
state_pre = new Matrix(DP, 1);
state_pre.Zero();
state_post = new Matrix(DP, 1);
state_post.Zero();
transition_matrix = new Matrix(DP, DP);
transition_matrix.SetIdentity(1.0);
transition_matrix[0, 1] = 1;
process_noise_cov = new Matrix(DP, DP);
process_noise_cov.SetIdentity(1.0);
measurement_matrix = new Matrix(MP, DP);
measurement_matrix.SetIdentity(1.0);
measurement_noise_cov = new Matrix(MP, MP);
measurement_noise_cov.SetIdentity(1.0);
error_cov_pre = new Matrix(DP, DP);
error_cov_post = new Matrix(DP, DP);
error_cov_post.SetIdentity(1);
gain = new Matrix(DP, MP);
if (CP > 0)
{
control_matrix = new Matrix(DP, CP);
control_matrix.Zero();
}
//
temp1 = new Matrix(DP, DP);
temp2 = new Matrix(MP, DP);
temp3 = new Matrix(MP, MP);
temp4 = new Matrix(MP, DP);
temp5 = new Matrix(MP, 1);
}
public Matrix Predict()
{
state_pre = transition_matrix.Multiply(state_post);
//if (CP>0)
//{
// control_matrix
//}
temp1 = transition_matrix.Multiply(error_cov_post);
Matrix at = transition_matrix.Transpose();
error_cov_pre = temp1.Multiply(at).Add(process_noise_cov);
Matrix result = new Matrix(state_pre);
return result;
}
public Matrix Correct(Matrix measurement)
{
temp2 = measurement_matrix.Multiply(error_cov_pre);
Matrix ht = measurement_matrix.Transpose();
temp3 = temp2.Multiply(ht).Add(measurement_noise_cov);
temp3.InvertSsgj();
temp4 = temp3.Multiply(temp2);
gain = temp4.Transpose();
temp5 = measurement.Subtract(measurement_matrix.Multiply(state_pre));
state_post = gain.Multiply(temp5).Add(state_pre);
error_cov_post = error_cov_pre.Subtract(gain.Multiply(temp2));
Matrix result = new Matrix(state_post);
return result;
}
public Matrix AutoPredict(Matrix measurement)
{
Matrix result = Predict();
Correct(measurement);
return result;
}
}
#endregion
public KalmanFacade(int valueItem)
{
if (valueItem<=0)
{
throw new Exception("not enough value items!");
}
kmfilter = new KalmanFilter[valueItem];
Random rand = new Random(1001);
for (int i = 0; i < valueItem; i++ )
{
kmfilter[i] = new KalmanFilter();
kmfilter[i].state_post[0, 0] = rand.Next(10);
kmfilter[i].state_post[1, 0] = rand.Next(10);
//
kmfilter[i].process_noise_cov.SetIdentity(1e-5);
kmfilter[i].measurement_noise_cov.SetIdentity(1e-1);
}
}
private KalmanFilter[] kmfilter = null;
public bool GetValue(double[] inValue, ref double[] outValue)
{
if (inValue.Length != kmfilter.Length || outValue.Length != kmfilter.Length)
{
return false;
}
Matrix[] measures = new Matrix[kmfilter.Length];
for (int i = 0; i < kmfilter.Length; i++ )
{
measures[i] = new Matrix();
measures[i][0, 0] = inValue[i];
outValue[i] = kmfilter[i].AutoPredict(measures[i])[0, 0];
}
return true;
}
}
}
//==========test=============
SimTransfer.KalmanFacade kalman = new SimTransfer.KalmanFacade(1);
Random rand = new Random(1001);
System.IO.StreamWriter dataFile = new System.IO.StreamWriter("D:\\test.csv");
for (int x = 0; x < 2000; x++ )
{
double y = 100 * Math.Sin((2.0 * Math.PI / (float)200) * x);
double noise = 20 * Math.Sin((40.0 * Math.PI / (float)200) * x) + 40 * (rand.NextDouble() - 0.5);
double[] z_k = new double[1];
z_k[0] = y + noise;
double[] y_k = new double[1];
kalman.GetValue(z_k, ref y_k);
dataFile.WriteLine(y.ToString() + "," + z_k[0].ToString() + "," + y_k[0].ToString());
}
dataFile.Close();
MessageBox.Show("OK!");
源码是很详细，但是注释比较少，看来我还得把程序翻译一遍！
这两天把注释重新写一下！！！
还有一个就是Matrix的类库。