Highly accurate pulse prediction for exercise equipment
In this thesis work, we propose a novel method for estimating a person's heart rate during exercise. There exists a commercial device used in treadmills and ellipticals with the same purpose. This commercial device is known to be slow and error prone in providing an estimate. We propose a method based on the optimal estimator for a statistical model of our design. We have tested our method on simulated data in MATLAB® as well as for six subjects. We have quantitative comparisons of our method against the commercially available POLAR® Heart Rate Board. On average our method was able to report an estimate 33% faster than POLAR with an order of magnitude less outliers. The method we developed could easily be implemented in real-time and marketed as a commercial alternative.^
Engineering, Electronics and Electrical
Andrew P O'Shea,
"Highly accurate pulse prediction for exercise equipment"
Dissertations and Master's Theses (Campus Access).