ABSTRACT
This chapter discusses the practical issues of coordination in control and optimization of large-scale stochastic systems and examines in what relation to them stands the existing theory of hierarchical control under uncertainty. It seems proper to begin with a discussion of iterative coordination under uncertainty. Taking into account the initial objectives related to system behavior, the investment and operating costs of the computing and communication facilities, and some appropriate quantitative measures of reliability, one could perhaps try to develop adequate mathematical models and to define in formal terms design problems related to the simultaneous choice of all structural and algorithmic decisions. Design problems involving the important issues discussed have not been formulated, and the limited theoretical research was confined to algorithmic design under very restricting assumptions, or to analysis of some chosen aspects of operation of hierarchical structures. At the opposite end of the "approach spectrum" one may place control structure with repetitive multirate forecast-based open-loop optimization.
