Neural network modeling of diagonal stride technique of highly qualified skiers with hearing impairments

Abstract

Abstract: Correct rational sport technique is based on models that are developed according to the kinematic characteristics of motion. According to the literature review, the questions addressed by the model are essential for studying the improvement of sport technique but the lack of those models for deaf athletes is a problem particularly in crosscountry skiing. So the purpose of this research is to develop a model of the kinematic structure of classical diagonal stride technique of highly qualified cross-country skiers with hearing impairments on the basis of computer neural networks. Participants: 9 high skilled skiers with hearing impairments of the Ukrainian National Deaflympic team on skiing. Results: The modeling kinematic indicators of the diagonal stride technique of highly qualified skiers with hearing impairments have been identified. Seven neural networks of multilayer perceptron type have been developed as a simulation of the velocity of the skier’s general center of mass in the movement cycle. On a basis of the best model, the errors in the diagonal stride technique of highly qualified skiers with hearing impairments were corrected. Conclusions: the neural network modeling in the process of technical performance improving of highly skilled skiers with hearing impairments is proved. Neural network modeling has allowed increasing the resultant velocity of skier’s general center of mass in the cycle of motion due to accounting skier’s individual biomechanical characteristics.