dynamic-cassie

This project computes minimum perturbation, rotational coordinates for OSU's Cassie robot. There are two project components: the MATLAB optimizer, and a python wrapper that does the data interpolation.

Python Wrapper

The optimal_cassie.py wrapper invokes the cassie.mat file, which contains samples, grid, and X, Y, Z optimal coordinates for Cassie. It defines several functions to load and use the coordinates.

load_data

Loads data stored in cassie.mat, and returns it in a dictionary.

interpolate_so3

Interpolation workhorse. Takes loaded data, a query point, and 3D rotation angle order, and returns an interp dictionary containing the resulting transform in a variety of forms.

interpolating_function

A simplified version of the above two functions. This returns a function, which takes just a query point and returns a 3D rotation matrix. This function can take the same arguments as the above two functions to change how the interpolating function is constructed.

MATLAB Optimizer

The optimizer uses the included 6dof_3lvl_sweep.txt data file to construct a local connection for Cassie. It then computes minimum perturbation coordinates, using a method published by Travers., et. al. All resulting data is saved in the included cassie.mat file; this should not need to be run except for higher resolution/higher DoF sweeps of the robot.

Note: The optimizer uses some functions from OSU-LRAM's sysplotter. A local installation is required to run the optimizer.