APPLICATION OF PHRASEOLOGICAL PATTERNS IN THE RECOGNITION OF CONTROLLER’S INSTRUCTIONS, FOR AIR TRAFFIC CONTROL TRAINING COMPLEXES

Abstract

The paper proposes an algorithm for recognizing ATC controller’s messages using method of contextual analysis of phraseology elements based on the application of phraseological patterns. A practical example of phraseological patterns for comprehensive recognition of ATC controller’s instruction is given, when a limited set of words included in this in- struction is known in advance. It is proved that the use of the recognition method with contextual analysis of the ATC phra- seology elements in conjunction with the recognition carried out by using software library intended for the universal recog- nition of the speech stream, will allow to increase the accuracy of the latter, while maintaining performance at a level suffi- cient to apply this method in control systems and real-time control. According to experts, the accuracy increase when using this method can be up to 30 %.