scholarly journals Estimation of Mechanical Power Output Employing Deep Learning on Inertial Measurement Data in Roller Ski Skating

Sensors ◽  
2021 ◽  
Vol 21 (19) ◽  
pp. 6500
Author(s):  
Md Zia Uddin ◽  
Trine M. Seeberg ◽  
Jan Kocbach ◽  
Anders E. Liverud ◽  
Victor Gonzalez ◽  
...  

The ability to optimize power generation in sports is imperative, both for understanding and balancing training load correctly, and for optimizing competition performance. In this paper, we aim to estimate mechanical power output by employing a time-sequential information-based deep Long Short-Term Memory (LSTM) neural network from multiple inertial measurement units (IMUs). Thirteen athletes conducted roller ski skating trials on a treadmill with varying incline and speed. The acceleration and gyroscope data collected with the IMUs were run through statistical feature processing, before being used by the deep learning model to estimate power output. The model was thereafter used for prediction of power from test data using two approaches. First, a user-dependent case was explored, reaching a power estimation within 3.5% error. Second, a user-independent case was developed, reaching an error of 11.6% for the power estimation. Finally, the LSTM model was compared to two other machine learning models and was found to be superior. In conclusion, the user-dependent model allows for precise estimation of roller skiing power output after training the model on data from each athlete. The user-independent model provides less accurate estimation; however, the accuracy may be sufficient for providing valuable information for recreational skiers.

Sports ◽  
2018 ◽  
Vol 6 (4) ◽  
pp. 151 ◽  
Author(s):  
Takafumi Kubo ◽  
Kuniaki Hirayama ◽  
Nobuhiro Nakamura ◽  
Mitsuru Higuchi

The aim of this study was to investigate whether accommodating elastic bands with barbell back squats (BSQ) increase muscular force during the deceleration subphase. Ten healthy men (mean ± standard deviation: Age: 23 ± 2 years; height: 170.5 ± 3.7 cm; mass: 66.7 ± 5.4 kg; and BSQ one repetition maximum (RM): 105 ± 23.1 kg; BSQ 1RM/body mass: 1.6 ± 0.3) were recruited for this study. The subjects performed band-resisted parallel BSQ (accommodating elastic bands each sides of barbell) with five band conditions in random order. The duration of the deceleration subphase, mean mechanical power, and the force and velocity during the acceleration and deceleration subphases were calculated. BSQ with elastic bands elicited greater mechanical power output, velocity, and force during the deceleration subphase, in contrast to that elicited with traditional free weight (p < 0.05). BSQ with elastic bands also elicited greater mechanical power output and velocity during the acceleration subphase. However, the force output during the acceleration subphase using an elastic band was lesser than that using a traditional free weight (p < 0.05). This study suggests that BSQ with elastic band elicit greater power output during the acceleration and deceleration subphases.


2021 ◽  
Vol 25 (11) ◽  
pp. 6041-6066
Author(s):  
Jiancong Chen ◽  
Baptiste Dafflon ◽  
Anh Phuong Tran ◽  
Nicola Falco ◽  
Susan S. Hubbard

Abstract. Climate change is reshaping vulnerable ecosystems, leading to uncertain effects on ecosystem dynamics, including evapotranspiration (ET) and ecosystem respiration (Reco). However, accurate estimation of ET and Reco still remains challenging at sparsely monitored watersheds, where data and field instrumentation are limited. In this study, we developed a hybrid predictive modeling approach (HPM) that integrates eddy covariance measurements, physically based model simulation results, meteorological forcings, and remote-sensing datasets to estimate ET and Reco in high space–time resolution. HPM relies on a deep learning algorithm and long short-term memory (LSTM) and requires only air temperature, precipitation, radiation, normalized difference vegetation index (NDVI), and soil temperature (when available) as input variables. We tested and validated HPM estimation results in different climate regions and developed four use cases to demonstrate the applicability and variability of HPM at various FLUXNET sites and Rocky Mountain SNOTEL sites in Western North America. To test the limitations and performance of the HPM approach in mountainous watersheds, an expanded use case focused on the East River Watershed, Colorado, USA. The results indicate HPM is capable of identifying complicated interactions among meteorological forcings, ET, and Reco variables, as well as providing reliable estimation of ET and Reco across relevant spatiotemporal scales, even in challenging mountainous systems. The study documents that HPM increases our capability to estimate ET and Reco and enhances process understanding at sparsely monitored watersheds.


2010 ◽  
Vol 628 (1-3) ◽  
pp. 116-127 ◽  
Author(s):  
Diethart Schmid ◽  
Dawid L. Staudacher ◽  
Christian A. Plass ◽  
Hans G. Loew ◽  
Eva Fritz ◽  
...  

2000 ◽  
Vol 203 (17) ◽  
pp. 2667-2689 ◽  
Author(s):  
R.K. Josephson ◽  
J.G. Malamud ◽  
D.R. Stokes

The basalar muscle of the beetle Cotinus mutabilis is a large, fibrillar flight muscle composed of approximately 90 fibers. The paired basalars together make up approximately one-third of the mass of the power muscles of flight. Changes in twitch force with changing stimulus intensity indicated that a basalar muscle is innervated by at least five excitatory axons and at least one inhibitory axon. The muscle is an asynchronous muscle; during normal oscillatory operation there is not a 1:1 relationship between muscle action potentials and contractions. During tethered flight, the wing-stroke frequency was approximately 80 Hz, and the action potential frequency in individual motor units was approximately 20 Hz. As in other asynchronous muscles that have been examined, the basalar is characterized by high passive tension, low tetanic force and long twitch duration. Mechanical power output from the basalar muscle during imposed, sinusoidal strain was measured by the work-loop technique. Work output varied with strain amplitude, strain frequency, the muscle length upon which the strain was superimposed, muscle temperature and stimulation frequency. When other variables were at optimal values, the optimal strain for work per cycle was approximately 5%, the optimal frequency for work per cycle approximately 50 Hz and the optimal frequency for mechanical power output 60–80 Hz. Optimal strain decreased with increasing cycle frequency and increased with muscle temperature. The curve relating work output and strain was narrow. At frequencies approximating those of flight, the width of the work versus strain curve, measured at half-maximal work, was 5% of the resting muscle length. The optimal muscle length for work output was shorter than that at which twitch and tetanic tension were maximal. Optimal muscle length decreased with increasing strain. The curve relating work output and muscle length, like that for work versus strain, was narrow, with a half-width of approximately 3 % at the normal flight frequency. Increasing the frequency with which the muscle was stimulated increased power output up to a plateau, reached at approximately 100 Hz stimulation frequency (at 35 degrees C). The low lift generated by animals during tethered flight is consistent with the low frequency of muscle action potentials in motor units of the wing muscles. The optimal oscillatory frequency for work per cycle increased with muscle temperature over the temperature range tested (25–40 degrees C). When cycle frequency was held constant, the work per cycle rose to an optimum with increasing temperature and then declined. We propose that there is a temperature optimum for work output because increasing temperature increases the shortening velocity of the muscle, which increases the rate of positive work output during shortening, but also decreases the durations of the stretch activation and shortening deactivation that underlie positive work output, the effect of temperature on shortening velocity being dominant at lower temperatures and the effect of temperature on the time course of activation and deactivation being dominant at higher temperatures. The average wing-stroke frequency during free flight was 94 Hz, and the thoracic temperature was 35 degrees C. The mechanical power output at the measured values of wing-stroke frequency and thoracic temperature during flight, and at optimal muscle length and strain, averaged 127 W kg(−1)muscle, with a maximum value of 200 W kg(−1). The power output from this asynchronous flight muscle was approximately twice that measured with similar techniques from synchronous flight muscle of insects, supporting the hypothesis that asynchronous operation has been favored by evolution in flight systems of different insect groups because it allows greater power output at the high contraction frequencies of flight.


2019 ◽  
Vol 14 (3) ◽  
pp. 303-309 ◽  
Author(s):  
Lotte L. Lintmeijer ◽  
A.J. “Knoek” van Soest ◽  
Freek S. Robbers ◽  
Mathijs J. Hofmijster ◽  
Peter J. Beek

2007 ◽  
Vol 39 (Supplement) ◽  
pp. S445-S446
Author(s):  
Pedro A. Galilea ◽  
Carlos González-Haro ◽  
Franchek Drobnic ◽  
Jesús F Escanero

2012 ◽  
Vol 113 (4) ◽  
pp. 584-594 ◽  
Author(s):  
Paola Zamparo ◽  
Ian L. Swaine

Determining the efficiency of a swimming stroke is difficult because different “efficiencies” can be computed based on the partitioning of mechanical power output (Ẇ) into its useful and nonuseful components, as well as because of the difficulties in measuring the forces that a swimmer can exert in water. In this paper, overall efficiency (ηO = ẆTOT/Ė, where ẆTOT is total mechanical power output, and Ė is overall metabolic power input) was calculated in 10 swimmers by means of a laboratory-based whole-body swimming ergometer, whereas propelling efficiency (ηP = ẆD/ẆTOT, where ẆD is the power to overcome drag) was estimated based on these values and on values of drag efficiency (ηD = ẆD/Ė): ηP = ηD/ηO. The values of ηD reported in the literature range from 0.03 to 0.09 (based on data for passive and active drag, respectively). ηO was 0.28 ± 0.01, and ηP was estimated to range from ∼0.10 (ηD = 0.03) to 0.35 (ηD = 0.09). Even if there are obvious limitations to exact simulation of the whole swimming stroke within the laboratory, these calculations suggest that the data reported in the literature for ηO are probably underestimated, because not all components of ẆTOT can be measured accurately in this environment. Similarly, our estimations of ηP suggest that the data reported in the literature are probably overestimated.


Sign in / Sign up

Export Citation Format

Share Document