Resumen
The subject matter of the article is ?V3-117 engine ratings and recognition methods. The goal of the work is to create methods for classification TV3-117 engine ratings based on neural network technologies in real time. The following tasks were solved in the article: the principles formation on classification and recognition of TV3-117 engine?s conditions, determination of main steps for solving problem of classification and recognition TV3-117 engine conditions in the neural network basis, development of a method for the classification and recognition TV3-117 engine conditions using neural networks. The following methods used are ? methods of probability theory and mathematical statistics, methods of neuroinformatics, methods of the information systems theory and data processing. The following results were obtained ? the principles of classification and recognition TV3-117 engine conditions are formulated and the main steps for solving this problem are defined. It is substantiated that solving the problem of classifying the TV3-117 engine ratings in the neural network basis allows solve this problem more efficiently with less time and computational resources than using classical methods (for example, the Bayes method). Conclusions: using the neural network technologies for the classification and recognition the TV3-117 engine conditions allows to reduce the processing time, and most of the time spent on solving this problem is used to train the neural network. Prospects for further research are the development of an expert system, one of the modules is the module of classification and recognition TV3-117 engine conditions which is used in the board system to monitor and diagnose the engine technical condition and interact with the engine control systems, allows is to effect to the executive mechanism fluently and in time, from the one hand, to improve the quality control engine and its subsystems from the other hand in order to increase its reliability during its operation.