过滤流数据以减少噪声,卡尔曼滤波c#
我正在将数据从惯性传感器传输到C#应用程序中。数据有点嘈杂,所以我需要添加一个过滤器来平滑它。我有一个kalman滤波器实现,当给定一个数组时效果很好,但我无法理解如何在恒定的数据流中使用它 我有:过滤流数据以减少噪声,卡尔曼滤波c#,c#,streaming,kalman-filter,C#,Streaming,Kalman Filter,我正在将数据从惯性传感器传输到C#应用程序中。数据有点嘈杂,所以我需要添加一个过滤器来平滑它。我有一个kalman滤波器实现,当给定一个数组时效果很好,但我无法理解如何在恒定的数据流中使用它 我有: double sensorData; //the noisy value, constantly updating from another class. 过滤器: public static double[] noisySine = new double[20] { 40, 41, 38, 40
double sensorData; //the noisy value, constantly updating from another class.
过滤器:
public static double[] noisySine = new double[20] { 40, 41, 38, 40, 45, 42, 43, 44, 40, 38, 44, 45, 40, 39, 37, 41, 42, 70, 44, 42 };
public static double[] clean = new double[20];
public static void KalmanFilter(double[] noisy)
{
double A = double.Parse("1"); //factor of real value to previous real value
// double B = 0; //factor of real value to real control signal
double H = double.Parse("1");
double P = double.Parse("0.1");
double Q = double.Parse("0.125"); //Process noise.
double R = double.Parse("1"); //assumed environment noise.
double K;
double z;
double x;
//assign to first measured value
x = noisy[0];
for (int i = 0; i < noisy.Length; i++)
{
//get current measured value
z = noisy[i];
//time update - prediction
x = A * x;
P = A * P * A + Q;
//measurement update - correction
K = P * H / (H * P * H + R);
x = x + K * (z - H * x);
P = (1 - K * H) * P;
//estimated value
clean[i] = x;
Console.WriteLine(noisy[i] + " " + clean[i]);
}
}
publicstaticdouble[]noisesine=newdouble[20]{40,41,38,40,45,42,43,44,40,38,44,45,40,39,37,41,42,70,44,42};
公共静态双精度[]干净=新双精度[20];
公共静态真空Kalman滤波器(双[]噪声)
{
double A=double.Parse(“1”);//实值到上一个实值的因子
//双B=0;//实际值对实际控制信号的系数
双H=double.Parse(“1”);
双P=double.Parse(“0.1”);
double Q=double.Parse(“0.125”);//进程噪声。
double R=double.Parse(“1”);//假定环境噪声。
双K;
双z;
双x;
//分配给第一个测量值
x=噪音[0];
for(int i=0;i
如何将一个double(而不是数组)流式输入并返回一个(过滤的)double
谢谢。试试下面的代码
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.IO;
namespace ConsoleApplication1
{
class Program
{
static void Main(string[] args)
{
double[] input = {1.1,2.2,3.3,4.4};
byte[] bArray = input.Select(x => BitConverter.GetBytes(x)).SelectMany(y => y).ToArray();
MemoryStream inStream = new MemoryStream(bArray);
long length = inStream.Length;
byte[] outArray = new byte[length];
inStream.Read(outArray, 0, (int)length);
List<double> output = new List<double>();
for (int i = 0; i < bArray.Length; i += 8)
{
output.Add(BitConverter.ToDouble(outArray,i));
}
}
}
}
使用系统;
使用System.Collections.Generic;
使用System.Linq;
使用系统文本;
使用System.IO;
命名空间控制台应用程序1
{
班级计划
{
静态void Main(字符串[]参数)
{
双[]输入={1.1,2.2,3.3,4.4};
byte[]bArray=input.Select(x=>BitConverter.GetBytes(x)).SelectMany(y=>y).ToArray();
MemoryStream inStream=新的MemoryStream(bArray);
长长度=河道内长度;
字节[]输出数组=新字节[长度];
流内读取(输出数组,0,(int)长度);
列表输出=新列表();
对于(int i=0;i
创建此类:
public class KalmanFilter
{
private double A, H, Q, R, P, x;
public KalmanFilter(double A, double H, double Q, double R, double initial_P, double initial_x)
{
this.A = A;
this.H = H;
this.Q = Q;
this.R = R;
this.P = initial_P;
this.x = initial_x;
}
public double Output(double input)
{
// time update - prediction
x = A * x;
P = A * P * A + Q;
// measurement update - correction
double K = P * H / (H * P * H + R);
x = x + K * (input - H * x);
P = (1 - K * H) * P;
return x;
}
}
并使用该类:
KalmanFilter filter = new KalmanFilter(1, 1, 0.125, 1, 0.1, noisySine[0]);
for (int i = 0; i < noisy.Length; i++) clean[i] = filter.Output(noisySine[i]);
KalmanFilter filter=新的KalmanFilter(1,1,0.125,1,0.1,noisesine[0]);
对于(inti=0;i
这就是如何修改代码以流式传输双精度输入,并返回过滤后的双精度输入
public static void KalmanTest()
{
double[] noisySine = new double[20] { 40, 41, 38, 40, 45, 42, 43, 44, 40, 38, 44, 45, 40, 39, 37, 41, 42, 70, 44, 42 };
for (int i = 0; i < noisySine.Length; i++)
{
Console.WriteLine(noisySine[i] + " " + KalmanFilter(noisySine[i]));
}
}
// assign default values
// for a new mwasurement, reset this values
public static double P = double.Parse("1"); // MUST be greater than 0
public static double clean = double.Parse("0"); // any value
public static double KalmanFilter(double noisy)
{
double A = double.Parse("1"); //factor of real value to previous real value
// double B = 0; //factor of real value to real control signal
double H = double.Parse("1");
double Q = double.Parse("0.125"); //Process noise.
double R = double.Parse("1"); //assumed environment noise.
double K;
double z;
double x;
//get current measured value
z = noisy;
//time update - prediction
x = A * clean;
P = A * P * A + Q;
//measurement update - correction
K = P * H / (H * P * H + R);
x = x + K * (z - H * x);
P = (1 - K * H) * P;
//estimated value
clean = x;
return clean;
}
publicstaticvoidkalmantest()
{
double[]noisesine=新的double[20]{40,41,38,40,45,42,43,44,40,38,44,45,40,39,37,41,42,70,44,42};
对于(int i=0;i
注意:有一个bug。当该代码迭代时,P很快成为接近R/100000的值,并且该行为与噪声无关,因为在P计算中没有对噪声或稳定读数的引用。
干净的代码看起来像一个低通滤波器:
// assign default values
public static double clean = double.Parse("0"); // any value
public static double KalmanFilter(double noisy)
{
double K = double.Parse("0.125"); // noise 0 < K < 1
clean = clean + K * (noisy - clean);
return clean;
}
//指定默认值
public static double clean=double.Parse(“0”);//任何价值
公共静态双Kalman滤波器(双噪声)
{
double K=double.Parse(“0.125”);//噪声0
double是八个字节。要流式传输数据,您需要一个字节数组。所以使用Bit.Converter类。嗨,谢谢你的回复。我不明白你的意思。我有一个变量(双)不断更新。我需要将其发送到一个过滤器函数中,该函数当前使用双[]。@anti您解决过这个问题吗?此实现中存在一个错误:当此代码迭代时,P很快变成接近R/100000的值,并且与噪声无关(在计算中没有对噪声或稳定读数的引用),谢谢,但我不认为这是我需要的。这如何帮助我将数据流发送到上面的过滤器函数中?如果我遗漏了什么,请道歉!我更新了代码以显示所有需要的代码。我使用了一个内存流。这不是一个bug,P(估计不确定性)为0表示错误