control.optimal.gaussian_likelihood_cost

control.optimal.gaussian_likelihood_cost(sys, Rv, Rw=None)[source]

Create cost function for Gaussian likelihoods

Returns a quadratic cost function that can be used for an optimal estimation problem. The cost function is of the form

cost = v^T R_v^{-1} v + w^T R_w^{-1} w

Parameters

sys (InputOutputSystem) – I/O system for which the cost function is being defined.
Rv (2D array_like) – Covariance matrix for input (or state) disturbances.
Rw (2D array_like) – Covariance matrix for sensor noise.

Returns

cost_fun – Function that can be used to evaluate the cost for a given disturbance and sensor noise. The call signature of the function is cost_fun(v, w).

Return type

callable