Development of a kick timing estimation algorithm in the crawl stroke for a biofeedback training system using neural oscillators

For the crawl stroke in swimming, it is important that the stroke made by the upper limbs and the flutter kick made by the lower limbs are well coordinated in order to enhance swimming performance. However, the training method to acquire the appropriate flutter kick timing has not been sufficiently...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Motomu NAKASHIMA, Takahiro MIYAZAWA, Yuji OHGI
Formato: article
Lenguaje:EN
Publicado: The Japan Society of Mechanical Engineers 2018
Materias:
Acceso en línea:https://doaj.org/article/8bd31e3b564a4ef4b0b03e25fdc75892
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
Descripción
Sumario:For the crawl stroke in swimming, it is important that the stroke made by the upper limbs and the flutter kick made by the lower limbs are well coordinated in order to enhance swimming performance. However, the training method to acquire the appropriate flutter kick timing has not been sufficiently established. In the present study, a biofeedback training system for swimmers to acquire appropriate kick timing was proposed. In this system, the most difficult and important part is the estimation of the appropriate kick timing. Therefore, the objective of this study was to develop a kick timing estimation algorithm in the crawl stroke for biofeedback training system using neural oscillators. First, a CPG network which outputs the kick estimation timing according to the input was constructed. In order to synchronize the roll angle in the CPG network with the actual one measured by a sensor, a special algorithm to change the cycle of the oscillation for the CPG network was introduced. Validation for the output of sinusoidal input accompanying sudden change in cycle was examined. It was found that the output signal for the roll tracked the input signal well, despite the sudden change in cycle. Validation for the actual input obtained in the experiment was next examined. It was found that the output from the CPG network was sufficiently consistent with the experimental values, suggesting sufficient performance of the proposed estimation algorithm.