Separation principle (source code)

= Separation principle
{wiki}

In a linear-quadratic-Gaussian control problem, the optimal control applies the full-information linear-quadratic feedback law to the conditional state estimate. The control gain is independent of the state-estimation error covariance.