A computational method for infinite-time optimal regulation problems

It is shown that, when an infinite-time optimal regulation problem is solved via a method of increasing the terminal time t
f
of a corresponding finite-time problem, convergence can be accelerated by incorporating a quadratic terminal weight
x(t_{f})^{T} Px(t_{f})
, where
P
is the steady-state solution of the Riccati equation for a corresponding truncated linear quadratic problem.