Restoration of partially lost heart rate data based on information about pedaling power in long-term cycling locomotion

ˑ: 

PhD E.D. Gorbunov1
PhD, Associate Professor A.V. Kubeev1
1Federal Science Center of Physical Culture and Sport (VNIIFK), Moscow

Objective of the study was to ensuring the completeness of accounting for physical activity performed by an athlete through the development and application of a technique that ensures the restoration of partially lost heart rate data based on information about pedaling power in long-term cycling locomotion.
Methods and structure of the study. The methodological basis of the developed technique was a mathematical model of the simplest neural network in the form of a single convolutional perceptron with a linear activation function. The perceptron was trained using available heart rate and pedaling power data using the least squares method. The methodology was tested on real data from a training recording with significant variability in heart rate.
Results and conclusions. A technique has been developed and tested that allows one to obtain estimates of heart rate based on pedaling power data. In the calculation example presented in the article, the mean square error in determining the heart rate relative to its true value was less than five beats per minute, which allows us to conclude that the results obtained are highly accurate.

Keywords: heart rate, pedaling power, perceptron, least squares method.

Работа выполнена в рамках государственного задания ФГБУ ФНЦ ВНИИФК № 777-00036-23-01 (код темы № 001-22/2).

References

  1. Alekseev E.R., Chesnokova O.V. Resheniye zadach vychislitelnoy matematiki v paketakh Mathcad 12, MATLAB 7, Maple 9 [Solving computational mathematics problems in Mathcad 12, MATLAB 7, Maple 9 packages]. Moscow: NT Press publ., 2006. 496 p.
  2. Venttsel E.S. Teoriya veroyatnostey [Probability theory]. Textbook for universities. 10th ed., ster. Moscow: Vysshaya shkola publ., 2006. 575 p.
  3. Volkov N.I., Oleynikov V.I. Bioenergetika sporta [Bioenergetics of sports]. Monograph. Moscow: Sovetskiy sport publ., 2011. 160 p.
  4. Nilsen E. Prakticheskiy analiz vremennykh ryadov: prognozirovaniye so statistikoy i mashinnoye obucheniye [Practical analysis of time series: forecasting with statistics and machine learning]. [transl.]. St. Petersburg: OOO «Dialektika» publ., 2021. 544 p.
  5. Rutkovskaya D., Pilinskiy M., Rutkovskiy L. Neyronnyye seti, geneticheskiye algoritmy i nechotkiye sistemy [Neural networks, genetic algorithms and fuzzy systems]. I.D. Rudinskiy [transl.]. Moscow: Goryachaya liniya – Telekom publ., 2006. 452 p.
  6. Yansen P. ChSS, laktat i trenirovki na vynoslivost [Heart rate, lactate and endurance training]. [transl.]. Murmansk: Tuloma publ., 2006. 160 p.