A Neuronet Navigational Training System
Abstract
Introduction: Training of aviation specialists is a time-consuming and expensive cyclical process requiring intermediate and final control. Purpose: An effective methodology should be developed for the organization of training of aviation specialists on the basis of axiological-competence approach with the use of training systems. The organization algorithm should cover both classroom and extracurricular learning in the study of special subjects. Results: Problems for a training system have been formulated: control, diagnostics, restoration of knowledge and skills at the theoretical and practical stages of the training, taking into account the adaptation to individual characteristics of the people working with the system (their level of training and psychophysiological features). It has been found that the design of training systems based on artificial neural networks allows you to keep the accuracy and reliability specified by the developer (without collecting and processing full statistical information about a group of trainees) when developing individual modules and entire systems. The learning process in training systems is considered a controlled procedure of solving adaptive tests with issuance of comments (diagnostics) and restoration of knowledge (compensation of missing or unassigned knowledge) by references to the theoretic material from an electronic training manual. An author's architecture has been proposed for a neural network navigation simulator and training system for the training of aviation specialists, which is based on the neural-network and neuron-fuzzy approach (using neural networks and fuzzy logic) to knowledge control; graphosemantic description of the subject area of the discipline under study; evaluation and derivation of each student's action when solving a problem based on binary trees followed by graphosemantic ranking of the complexity of individual operations, as well as the analysis of how close an answer is to the correct one. The architecture makes it possible to provide variability (the learners can choose the order in which they study the material) and adapt the learning process to the individual traits of the trainees. Practical relevance: The developed system allows you to reduce the training time, to intensify the learning process, to improve mastering of knowledge, and to perform remote monitoring at all stages of the training.Published
2017-06-21
How to Cite
Grigoryev, A., & Burlutsky, S. (2017). A Neuronet Navigational Training System. Information and Control Systems, (3), 89-98. https://doi.org/10.15217/issn1684-8853.2017.3.89
Issue
Section
Information instrumentation and education