Incomplete decomposition algorithms for discrete dynamic nonlinear models

The philosophy of incomplete decomposition algorithms is discussed where the system is considered as a single unity at each iteration. The algorithm is free to use any method to improve the approximation at each partial iteration. The method works as if the ith problem were a real life problem for which there exists a specific team dedicated to its solution, probably with different tools. In this way, each team takes advantage of the progress of the team that are its predecessors, minimizing the inactive time.