KalmanFilter Constructor

public KalmanFilter(
	double processNoise = 1E-05,
	double measurementNoise = 0.001,
	double estimatedError = 1E-05
)

Public Sub New ( 
	Optional processNoise As Double = 1E-05,
	Optional measurementNoise As Double = 0.001,
	Optional estimatedError As Double = 1E-05
)

public:
KalmanFilter(
	double processNoise = 1E-05, 
	double measurementNoise = 0.001, 
	double estimatedError = 1E-05
)

new : 
        ?processNoise : float * 
        ?measurementNoise : float * 
        ?estimatedError : float 
(* Defaults:
        let _processNoise = defaultArg processNoise 1E-05
        let _measurementNoise = defaultArg measurementNoise 0.001
        let _estimatedError = defaultArg estimatedError 1E-05
*)
-> KalmanFilter

GSF.NumericalAnalysis.KalmanFilter = function(processNoise, measurementNoise, estimatedError);

Parameters

processNoise  Double  (Optional)

Determines how much the system state is expected to change between measurements.

Start with a very small value (e.g., 1e-5) and gradually increase it. Too small a value can make the filter slow to adapt to changes, while too large can make it over-responsive to noise.

measurementNoise  Double  (Optional)

Reflects the confidence in the measurements. A lower value gives more weight to the measurements.

If your measurements are accurate, set R to a small value (e.g., 1e-3). Increase it if the measurements are noisy.

estimatedError  Double  (Optional)

Represents the initial guess about the error in the state estimate.

Start with a value that reflects the expected variability in the initial state.

Reference