Fuzzy Estimation of VO2 Dynamics During Cycling Exercise

Abstract

One of the most important variables in cycling is the oxygen consumption (VO2) and its maximal value (‘VO2max’). The latter is considered as an appropriate indicator of the cardio-respiratory fitness and provides interesting information for many applications in the sports medicine context, however, this variable is not easy to measure out of a laboratory. Hence, this paper presents an alternative approach to the estimation of VO2 dynamics by means of fuzzy systems that use an input vector composed of three easy-to-obtain variables: Heart Rate, Work Rate, and Respiratory Rate measured in a clinical Cardiopulmonary-Exercise Testing. Two tuning strategies are compared: the well-known adaptive-network-based fuzzy inference system and an evolutionary fuzzy system based on the differential evolution algorithm. Experimental results showed that both tuning strategies are capable of providing competitive solutions in terms of several regression indices.