The work shown here was done as part of my Ph.D. degree at the Mechanical and Aerospace Engineering Department of the University of California, Irvine. I started working with Eigenspostures sometime during 2012 and it became an integral part of my Ph.D. Thesis.
The main goal of this project was to create an efficient algorithm to generate and analyze high-diving motions, with the idea to extend it to other complex multibody systems. Usually, to generate diving motions (or any other complex motion for multibody systems) one would set up and solve an optimal control problem. However, as the complexity of the system increases, so do the complexity of the solutions.
See the following video, for a few examples of diving motions:
The main idea was to use “Eigenpostures” to simplify the optimal control problem. Eigenpostures have a lot in common with mode shapes and are used in a very similar way. By using Eigenpostures, we were able to efficiently generate diving motions. For instance:
or
The use of Eigenpostures reduced the number of degrees of freedom by up to two orders of magnitude and made setting up and solving the optimal control problem faster.
A second good use of Eigenpostures comes from motion analysis. One can map existing dives to an “Eigenposture space” to intuitively obtain information about the dive or to try to understand which features in the dive are more relevant to the score. This feature can be of great interest in a variety competitive sports, providing a tool to identify which traits in the motion improve performance the most.
For more information about the work discussed here, see Part I of my Ph. D. Thesis. If you are interested in the technicalities (code) of the work discussed here, please send me an email at joana1@uci.edu.