Real-time track cycling performance prediction using ANFIS system

ABSTRACT

The next stage performance evaluation of an athlete can be predicted by implementing Artificial Intelligence technique. In track cycling event, coach and sports physician are concerned with the performance of the cyclist. The performance prediction may help to fine-tune the cyclist training intensities and strategies planning. This study was conducted to fulfil the prediction requirement by adopting a Fuzzy Inference System to classify the cyclist current cycling performance state. The six levels of output classification by a Fuzzy Inference System are to indicate the athlete’s current state performance using the body temperature, heart rate variability and speed as input parameters. An Adaptive Neuro-Fuzzy Inference System was applied to predict the cycling speed that can be achieved in the next lap. Using Adaptive Neuro-Fuzzy Inference System method, the average speed for the next laps can be predicted and compared with the actual speed. The regression value with r = 0.9029 indicates the Adaptive Neuro-Fuzzy Inference System is an adequate prediction algorithm to evaluate the cyclist performance. The predicted time to complete compared favourably with the actual finishing time with a ± 13.6% average error. Hence, the developed system is reliable and suitable for sports events that deal with speed and time.

KEYWORDS:

• Cyclist performance prediction
• fuzzy inference system
• adaptive neuro-fuzzy inference system
• real-time prediction
• performance classification

Acknowledgments

This research is funded by the Sports Division, Ministry of Higher Education Malaysia under the Sports Research Grant Scheme (Grant number: SRGS 9019-00002) and Knowledge Transfer Program (Grant number: KTP 9011-00023). This research is also supported by the National Sports Council (MSN), National Sports Institute (ISN) and Malaysia Junior Track Cycling Team.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Ministry of Higher Education, Malaysia [SRGS 9019-00002]; [KTP 9011-00023].