Package org.hipparchus.filtering.kalman.extended

Interface NonLinearProcess<T extends Measurement>

Type Parameters:

T - the type of the measurements

public interface NonLinearProcess<T extends Measurement>

Non-linear process that can be estimated by a ExtendedKalmanFilter.

This interface must be implemented by users to represent the behavior of the process to be estimated

Since:

1.3

See Also:

• ExtendedKalmanFilter
• LinearProcess

Method Summary

Modifier and Type	Method	Description
NonLinearEvolution	getEvolution(double previousTime, RealVector previousState, T measurement)	Get the state evolution between two times.
RealVector	getInnovation(T measurement, NonLinearEvolution evolution, RealMatrix innovationCovarianceMatrix)	Get the innovation brought by a measurement.

Method Details

getEvolution

NonLinearEvolution getEvolution(double previousTime,
 RealVector previousState,
 T measurement)

Get the state evolution between two times.

Parameters:

previousTime - time of the previous state

previousState - process state at previousTime

measurement - measurement to process

Returns:

state evolution

getInnovation

RealVector getInnovation(T measurement,
 NonLinearEvolution evolution,
 RealMatrix innovationCovarianceMatrix)

Get the innovation brought by a measurement.

Parameters:

measurement - measurement to process

evolution - evolution returned by a previous call to getEvolution(double, RealVector, Measurement)

innovationCovarianceMatrix - innovation covariance matrix, defined as $h.P.h^T+r$ where h is the measurement Jacobian, P is the predicted covariance and r is measurement covariance

Returns:

innovation brought by a measurement, may be null if measurement should be rejected