A Model of Human Decision Making in Complex Systems and its Use for Design of System Control Strategies

The paper describes a model of operators' decision making in complex system control, based on studies of event reports and performance in control rooms. This study shows how operators base their decisions on knowledge of system properties at different levels of abstraction depending on their perception of the system's immediate control requirements. These levels correspond to the abstraction hierarchy including system purpose, functions, and physical details, which is generally used to describe a formal design process. In emergency situations the task of the operator is to design a suitable control strategy for systems recovery, and the control systems designer should provide a man-machine interface, supporting the operator in identification of his task and in communication with the system at the level of abstraction corresponding to the immediate control requirement. A formalized representation of system properties in a multilevel flow model is described to provide a basis for an integrated control system design.