Cognitive control systems based on DIKW-pyramids of knowledge and multi-level subsystems of activity that use the dynamic complex of the goals of their functioning are considered. Describes the characteristics of individual goals and sets of these objectives, such as type, coverage, clarity of setting, income from achieving the goal, the need for system resources, priority, the degree of influence on other goals, competing system resources, the effectiveness of goals, and others. These characteristics determine the state of the processes of achieving the goal in the cognitive control system.
The activity of the control system for the selection of a strategy for managing goals is allocated to a separate sub-level in the pyramid of cognitive control system activities. The principles of choosing a goal-setting strategy based on the goal tree traversal, expansion of the core and the elimination of ineffective goals are formulated. The block diagram of the algorithm for selecting one goal is given. A fuzzy controller for target selection is described. Functional structures are proposed for the implementation of the goal control, which represent the upper levels in the subsystem of the cognitive control system. These levels consist of input and output operating machines, as well as finite control automata. The input operating automata handle the knowledge that comes from the knowledge and cognition subsystems, and the output operating automata form the knowledge for these subsystems. At the level of goal control, the FSM initiate the selection of a scenario for achieving the goals. In turn, these state machines are initiated and adapted by signals from the goal-setting level.