Using response surface models to maximize performance in swimming
Presenter: Wei-han Tan & Wei-fu Tan
Northside Health Careers High School, San Antonio, TX, USA
摘要：
The researchers are interested in whether statistics and computers can be used to maximize a swimmer's performance in 200 yard races. In this initial year of studying swimming, if statistics and computers can benefit the swimmer, evidence will be seen. Ultimately, the researchers hypothesize that it is possible for this methodology to benefit Swimming.
The initial 50 yard split time, second 50 yard split time and third 50 yard split time were used as a function of a swimmer's racing characteristics. Each swimmer has their own optimum combination of these variables to perform at their best.
To begin the experiment, judgment modeling is used. Each swimmer begins with identical judgment models that bound their swimming characteristics, ensuring that their optimum combination be found within the range of data. After the first swim, each individual's data is shifted toward their actual swimming characteristics. After every input of data, the swimmer's optimum combination of the three variables is computed. As the swimmers produce data, the data is combined with the previous data in process known as evolutionary operations. The adjustments in the model/optimum values are based upon the swimmer's physical improvement or adjustment in swimming tactic.
It has been concluded that after further research, this methodology can be used to help guide swimmers in maximizing their performance. Promising results were found, because the three measured variables of initial 50 yard split time, second 50 yard split time, and third 50 yard split time were all found to be significant.