The PMSS approach to large-scale optimal control problems (Q798277)

The authors take the view that there are two aspects to solving nonlinear programming problems, namely, problem manipulation (PM) and solution strategy (SS). An example of the former is the construction of the dual of a linear program to obtain a simpler problem, while an example of the latter is replacing a constrained problem by a sequence of unconstrained problems using a penalty function. Using these two concepts (PM and SS) an attempt is made to show how various algorithms in decentralized large scale control problems can be put into a systematic framework.