For a long time, baseball lovers have been interested in the physics of a baseball’s flight. One particularly big question is how the seams on a baseball affect the way it flies from the pitcher to the home plate. A mechanical engineer, LeRoy Alaways has been pursuing the answer to this question. Identifying factors such as the speed and orientation of a ball’s spin as well as its velocity and location at release, Alaways has created a system of equations using these factors to predict pitch trajectories. But to know that, Alaway needed to know the initial conditions of the pitch.
So, Alaway recorded the ball trajectories on 10 high-speed video cameras, and using non-linear least squares parameter estimation, he worked backwards to find out the acceleration and spin put on the ball by the pitcher. These parameters are plugged into Alaway’s system of equations and adjusted by a computer to best match the trajectories produced by the video.
Alaway’s research has led to a better understanding of how seams affect pitches - apparently the two different ways of pitching become less relevant “as the speed at which the seams circulate begins to compare to the speed of the ball itself”. The research has also led to the creation of a new pitching machine. The machine takes into account Alaway’s equations and can perform any kind of pitch, accurately putting the ball “within a palm’s width of the strike zone.”
This story exemplifies the value of least-squares methods to the outside world. The method used by Alaway to work backwards from the ball trajectories to the initial parameters reminded me of project 2. The trajectories would be the huge “A” matrix, and the initial parameters would be the light source “x” we wanted to find. This is not to say that non-linear least squares parameter estimation is like SVD, but rather that doing project 2 helped me better understand how Alaway might have used least-squares to get his initial parameters, which, to me, is another example of a connection CS 322 has to the real world.
http://findarticles.com/p/articles/mi_m1200/is_23_159/ai_76463830/pg_1






Leave a Comment
You must be logged in to post a comment.
* You can follow any responses to this entry through the RSS 2.0 feed.