Nonlinear approximations for the solution of team optimal control problems
Marco Baglietto, Thomas Parisini, R. Zoppoli · 2002
The method of constraining unknown control functions to take the structures of suitable nonlinear approximators enables one to solve approximately (but at any desired degree of accuracy and without the need of too many parameters to optimize) "difficult" functional optimization problems. The method is tested on Witsenhausen's (1968) counterexample.