Neuro-fuzzy control system of a mobile plant using the brain computer interface

The complexity of control systems grows with the growth of the complexity of the control plant. Incompleteness, conflicting source data for control system design (development) does not allow constructing the ideal control laws. As a result, the synthesized control laws do not exclude the possibility of the deviation from the system’s target functioning. Maximum deviations lead to emergencies. Numerous methods and algorithms of adaptation, self-adjustment, self-organization, in the majority are algorithmically difficult. As a result, processes of adaptation proceed rather slowly. Therefore, adaptive, self-organizing control systems are making to control slow-changing processes. In critical situations, they cannot provide accomplishment of target requirements of accuracy, quality, and high-speed performance of control process. Most often, the mode of a protective shutdown is provided in emergencies. However, for example, for mobile objects (aircraft, land/surface/underwater vehicles) switching off the control system will inevitably lead to an accident. Therefore, the systems of supervisory control are widely applied to the control of critical objects. The operator of the supervisory control system observes the work of an automatic system and reacts to the emergence of deviations from the target behavior of the control plant. In an emergency, the automatic transition to the mode of “manual” management is provided. However, the operator cannot always timely manage to evaluate a situation, to make a decision, and to give a command (for operation execution) for prevention of accident. Especially as on execution of the command of the operator by operating mechanism some time is also required. Neurophysiologic researches of electroencephalograms data (EEG-data) allow detecting characteristic fragments (patterns) corresponding to certain types of brain activity. Some of such patterns can be interpreted as the moments of decision-making by the operator. At the same time manifestation of “the decision-making pattern” precedes the manifestation of “the pattern of accomplishment of actions”. The precedence of “the decision-making pattern” before “the pattern of accomplishment of actions” can reach several minutes. Timely detection of “the decision-making pattern” of the operator’s brain activity the initiation of execution of the actions corresponding to “the pattern of accomplishment of actions” of the operator is possible. Our work is devoted to the research of ways of control systems design of the mobile object with the use of the analysis of brain activity of the operator (brain-computer interface).