RU94030327A - Method for recognition of state of complex objects - Google Patents

Abstract

FIELD: technical cybernetics. SUBSTANCE: method may be applied for automatic control systems which solve tasks of recognition and estimation of states of complex objects. Method provides composite estimation of tolerant space with multiple dimensions using remote control data. Novel feature of method is integral estimation of fact that objects transits from some class of states to another one. Estimation uses integral criteria that are generalized by set of parameters. The transition is detected by means of way how intersection area of multiple-parameter tolerant space between said classes of transition. This is represented by ordered set of tolerant sub-spaces, each of which correspond to remote-measured parameter and are ordered by error value. Said parameters provide information about object state and are encoded for each class of object states by some color in visible range. Set of all tolerant sub- spaces when object transits from one class of states to another one is marked by special color and is ordered by error value. Method provides possibility to design efficient real-time expert systems for estimation of state of complex objects. EFFECT: increased speed, increased precision of recognition.

Claims (1)

The method for recognizing the states of complex objects belongs to the field of technical cybernetics and can be used in automated control systems to solve problems of recognizing and evaluating the states of complex objects. The aim of the invention is to increase the speed and accuracy of recognition of the states of complex objects. This is achieved through a comprehensive assessment of the multi-parameter space of tolerance between classes of object states according to telemetric control data. New in the proposed method for recognizing the states of complex objects is an integral, generalized in terms of a number of parameters, assessment of the fact of the transition of an object from one class of states to another by the nature of the representation of the overlap region of the multiparameter tolerance space between these classes in the form of an error-ordered set of tolerance subspaces for each telemetered parameter, carrying information about the state of the object and encoded for each class of state of the object defined the color code of the scale of the visible spectrum, with a separate color highlighting the aggregate of all tolerance subspaces ordered by error at the time of transition of one class of object states to another. Based on the use of the present invention, highly efficient automated expert systems for assessing the state of complex objects operating in real time can be created.