Typedef lupnt::FilterMeasurementFunction

Typedef Documentation

typedef std::function<VecXd(const State &x, MatXd *H, MatXd *R)> lupnt::FilterMeasurementFunction

Function signature for predicting a measurement from the current state, and computing its Jacobian H and noise covariance R.

Stored in Filter::f_meas_ (via SetMeasurementFunction) and called once per Update step by EKF::Update/UDUEKF::Update (and once per sigma point plus once for sizing by UKF::Update) to form the measurement residual dz = z_true - z_prior. Typically produced from a plain MeasurementFunction via GetFilterMeasurementFunction, and often built from a measurement model’s CreateFunction (e.g. GnssMeasurement::CreateFunction).

Param x:

Current state estimate

Param H:

Output measurement Jacobian d(z)/d(x), size [n_z x n_x]

Param R:

Output measurement noise covariance, size [n_z x n_z]

Return:

Predicted measurement vector, size [n_z]