The paper presents the basic principles and main properties of a new modified feasible method that can be applied in the steady-state optimisation of complex systems. The essential feature of the proposed structure is that the procedures for finding the optimal solution for the steady-state optimisation problems, on both the coordinator level and on the local one, consist of solving linear algebraic equations. This is important from a practical point of view. Two examples of optimal steady-state control of a complex system are given.