The analysis and utilization of cycling training data.

Area of Science:

• Sports Science
• Exercise Physiology
• Biomathematics

Background:

• Accurate quantification of training intensity and quantity (dose) is essential for mathematical modeling of athletic training and performance.
• Traditional methods like training volume (e.g., km/week) often neglect training intensity's impact.
• Emerging methods offer more comprehensive quantification of training stimulus.

Purpose of the Study:

• To summarize available methods for quantifying training dose, particularly in cycle sport.
• To review mathematical techniques for modeling the relationship between training and performance.
• To evaluate the applicability of various quantification and modeling approaches.

Main Methods:

• Review of scientific literature on training quantification and performance modeling.
• Description of subjective (session rating of perceived exertion) and objective (heart rate-derived training impulse - TRIMP) quantification methods.
• Discussion of mathematical modeling approaches including impulse-response models, artificial neural networks, and other advanced techniques (PerPot, mixed linear modeling, chaos theory).

Main Results:

• Training volume alone is insufficient; intensity must be considered.
• TRIMP and normalized power (for cycling) are valuable metrics for quantifying training dose and physiological response.
• Impulse-response models offer insights but require extensive data; artificial neural networks and other advanced methods show potential but need further research.

Conclusions:

• Effective athletic training modeling requires robust quantification of both training intensity and quantity.
• Advanced methods like normalized power in cycling and TRIMP provide more accurate training dose assessments.
• While traditional models have limitations, ongoing research into systems theory, neural networks, and other mathematical approaches promises improved understanding of training-performance dynamics.