static void Main(string[] args)
{
int T = 1000;
double dt = 1;
Matrix<double> trajectory = Matrix<double>.Build.Dense(3, T, 0);
for (int t = 0; t < T; t++)
{
trajectory[0, t] = Math.Cos(2*Math.PI*t/T);
trajectory[1, t] = Math.Sin(2 * Math.PI * t / T);
trajectory[2, t] = 0;
}
string name = DateTime.Now.ToString("dd_MM_yyyy_hh_mm_ss");
MatlabWriter.Write<double>(name + "_clean" + ".mat", trajectory, "X1");
trajectory += Matrix<double>.Build.Random(3, T)/10;
MatlabWriter.Write<double>(name + "_raw" + ".mat", trajectory, "X2");
Vector<double> x = Vector<double>.Build.Dense(9, 0);
//x.SetSubVector(0, 3, trajectory.Column(0));
double p = 0.00001;
double q = 0.00000001;
double r = 1;
KfParameters parameters = new KfParameters(9, 3, x, dt, p, q, r);
KalmanFilter kalmanfilter = new KalmanFilter();
kalmanfilter.InitialiseKalmanFilter(parameters);
for (int t = 0; t < T; t++)
{
kalmanfilter.Forward(trajectory.Column(t));
}
MatlabWriter.Write<double>(name + "_filtered" + ".mat", kalmanfilter._kfParameters.X, "X3");
}
/// <summary>
/// Initialise optimal Kalman Filter
/// </summary>
/// <param name="kfParameters">Parameter object for this Kalman Filter</param>
public void InitialiseKalmanFilter(KfParameters kfParameters)
{
this._kfParameters = kfParameters;
}